{ "schema": "postcolor-public-listening-manifest-v1", "audio_count": 9, "new_audio_count": 9, "reused_reference_count": 27, "selection": "RAW_POSTCOLOR_M1のみ。5録音のnative 0半音/seed0を5本、sample3（121-121726-0012）の±3半音/seed 0・20261015を4本。", "unselected_waveforms_included": false, "source_NPY_files_included": false, "conversion": "保存済みfloat64 NPYから先頭min(N,80000)標本を抜粋。gain 1、無正規化、rint(x*32767)のnearest ties-to-even、little-endian PCM16、mono、16 kHz、無dither。clip・再標本化・時刻整列なし。", "interpretation": [ "この変換は試聴用の形式変換であり、新合成・品質測定ではない。", "波形・bytesの一致や元音声への近さだけでは品質が良いと判断しない。", "LibriSpeech原音は内容・時刻の参考であり、音高編集音の正解音声ではない。", "全長の評価結果と最大5秒の公開試聴例は区別する。", "RAW_POSTCOLOR_M1は固定した比較方式であり、採用済みの開発ベースとはしない。" ], "source_integrity": { "final_delivery_receipt": { "bytes": 48967, "sha256": "56cfc77a7ee958570970ef52ab9c38df5bb45b8ef9a92f3af94464059afa9dc9" }, "all_30_waveforms_csv": { "bytes": 10096, "sha256": "e49c4bd59d6e7503a007225aef60652606366f66bb439a25e24248ad938582fc" }, "all_nine_source_file_hashes_and_bytes_matched": true, "baseline_PML_and_RAW_FULL_condition_match_count": 18 }, "existing_public_manifest": { "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio-manifest.json", "bytes": 27563, "sha256": "19e37f3f3d024ff5b48c9f2bcb6b4397a6573de145bfea6d7fdff4cc4556e0f1" }, "existing_public_report_url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/index.html", "attribution": { "dataset": "LibriSpeech ASR Corpus / OpenSLR SLR12", "copyright_notice": "LibriSpeech (c) 2014 Vassil Panayotov", "authors": [ "Vassil Panayotov", "Guoguo Chen", "Daniel Povey", "Sanjeev Khudanpur" ], "source_url": "https://www.openslr.org/12/", "license": "Creative Commons Attribution 4.0 International (CC BY 4.0)", "license_url": "https://creativecommons.org/licenses/by/4.0/", "license_legalcode_url": "https://creativecommons.org/licenses/by/4.0/legalcode", "citation": "V. Panayotov, G. Chen, D. Povey, S. Khudanpur. LibriSpeech: An ASR Corpus Based on Public Domain Audio Books. ICASSP, 2015.", "paper_url": "https://www.danielpovey.com/files/2015_icassp_librispeech.pdf", "upstream_sources": "LibriVox audiobook recordings and Project Gutenberg texts", "recording_ids": [ "1089-134686-0008", "1188-133604-0003", "121-121726-0012", "1221-135766-0005", "1089-134686-0032" ], "research_modifications": "Fixed source excerpts up to six seconds, PML resynthesis and unseparated high-band envelope transfer. Public listening excerpts retain existing PCM samples, with only first-five-second cropping for longer recordings.", "endorsement": "No endorsement by the authors, distributors, readers, or source projects is implied." }, "additional_research_modifications": "RAW_POSTCOLOR_M1の保存済み再合成波形を、無正規化・gain1でPCM16へ量子化。長い録音は先頭5秒に抜粋。音高編集条件は±3半音。元の全著者・著作権表示・出典・CC BY 4.0を維持する。", "audio": [ { "case": "librispeech_1089-134686-0008", "sample": 1, "recording_id": "1089-134686-0008", "pitch_semitones": 0, "seed": 0, "variant": "RAW_POSTCOLOR_M1", "file": "audio/1089-134686-0008__pitch+0__seed0__RAW_POSTCOLOR_M1.wav", "source_NPY_sha256": "3fb4ef0208790e1d9a08e144860ff35699c058e53a100462576d2842e37c965f", "source_NPY_file_sha256": "3fb4ef0208790e1d9a08e144860ff35699c058e53a100462576d2842e37c965f", "source_NPY_bytes": 768128, "source_array_sha256": "8798be304887ebf36ab54582dec53c1fcb3c3c145ea605d1c12b209170bf2ce8", "source_array_sha256_definition": "NPYヘッダーを除く全長のlittle-endian float64標本列。C-orderの連続bytes。", "source_array_bytes": 768000, "source_dtype": "float64", "source_byte_order": "little", "source_samples": 96000, "source_sample_rate": 16000, "source_all_finite": true, "source_nonfinite_count": 0, "source_min": -0.4564454418141395, "source_max": 0.40033120795851573, "source_peak": 0.4564454418141395, "samples": 80000, "sample_rate": 16000, "channels": 1, "sample_width_bytes": 2, "excerpt_sample_range": [ 0, 80000 ], "excerpt_sample_range_end": "exclusive", "excerpt_seconds": [ 0, 5.0 ], "excerpt_peak": 0.4564454418141395, "cut_reason": "先頭5秒まで", "gain": 1, "normalization": "none", "dither": "none", "resampling": "none", "alignment": "none", "quantization": "rint(x*32767), nearest ties-to-even, little-endian signed PCM16", "clipping": "none", "clip_required": false, "clip_count": 0, "PCM_min": -14956, "PCM_max": 13118, "source_file_identity_matched": { "final_receipt": true, "all_30_waveforms_csv": true, "condition_measurement": true }, "baseline_comparisons": [ { "variant": "PML", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/1089-134686-0008__pitch+0__seed0__PML.wav", "WAV_sha256": "54d2b141b5b7e67bc735ca5e399f54a935853956e8f5ff7286b1651865f71d51", "bytes": 160044, "source_NPY_sha256": "66fa9fe6bb6e0880a2213e700ec59f6dd8af48283c5c350d980d948bf12644f6", "condition_and_source_NPY_identity_matched": true, "excerpt_sample_range": [ 0, 80000 ] }, { "variant": "RAW_FULL", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/1089-134686-0008__pitch+0__seed0__RAW_FULL.wav", "WAV_sha256": "eb7822699fd9ed0aeb2382a2edb2a3a358541787286b74f46eaceb01add4b9e1", "bytes": 160044, "source_NPY_sha256": "64a0d355df20f0bc11c6826f22412a88cb5259f79a7f1b115ad836c18fe43376", "condition_and_source_NPY_identity_matched": true, "excerpt_sample_range": [ 0, 80000 ] } ], "bytes": 160044, "WAV_sha256": "f2820245982df2dc9ecbc6089331029bee58b7a7e1789a8ded4e0b5e6596ad1f", "header": { "container": "RIFF/WAVE", "format_code": 1, "format": "PCM", "endianness": "little", "channels": 1, "sample_rate": 16000, "byte_rate": 32000, "block_align": 2, "bits_per_sample": 16, "sample_width_bytes": 2, "samples": 80000, "header_bytes": 44, "riff_size": 160036, "data_bytes": 160000, "header_hex": "524946462471020057415645666d74201000000001000100803e0000007d0000020010006461746100710200", "header_sha256": "fe37d7895050fe71672e015eccb2ba6a4ca09fed7bdcd9a838adf2c905a2ca42", "PCM_data_sha256": "e59e03a7bcd329bb0c41410846441adad343783360ee4ac8eef1c7304e5265db" }, "written_PCM_matches_declared_conversion": true }, { "case": "librispeech_1188-133604-0003", "sample": 2, "recording_id": "1188-133604-0003", "pitch_semitones": 0, "seed": 0, "variant": "RAW_POSTCOLOR_M1", "file": "audio/1188-133604-0003__pitch+0__seed0__RAW_POSTCOLOR_M1.wav", "source_NPY_sha256": "35d484a333cf1dab94d3febcb9d0d2105a85804fe4db39537a8f57a5e5843fb3", "source_NPY_file_sha256": "35d484a333cf1dab94d3febcb9d0d2105a85804fe4db39537a8f57a5e5843fb3", "source_NPY_bytes": 768128, "source_array_sha256": "cfcdc013025bda81e6a34cd5a1920c5d2ddd03456be325f9f8ba8c422429f92a", "source_array_sha256_definition": "NPYヘッダーを除く全長のlittle-endian float64標本列。C-orderの連続bytes。", "source_array_bytes": 768000, "source_dtype": "float64", "source_byte_order": "little", "source_samples": 96000, "source_sample_rate": 16000, "source_all_finite": true, "source_nonfinite_count": 0, "source_min": -0.3527812415268272, "source_max": 0.22094914223634987, "source_peak": 0.3527812415268272, "samples": 80000, "sample_rate": 16000, "channels": 1, "sample_width_bytes": 2, "excerpt_sample_range": [ 0, 80000 ], "excerpt_sample_range_end": "exclusive", "excerpt_seconds": [ 0, 5.0 ], "excerpt_peak": 0.3527812415268272, "cut_reason": "先頭5秒まで", "gain": 1, "normalization": "none", "dither": "none", "resampling": "none", "alignment": "none", "quantization": "rint(x*32767), nearest ties-to-even, little-endian signed PCM16", "clipping": "none", "clip_required": false, "clip_count": 0, "PCM_min": -11560, "PCM_max": 7240, "source_file_identity_matched": { "final_receipt": true, "all_30_waveforms_csv": true, "condition_measurement": true }, "baseline_comparisons": [ { "variant": "PML", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/1188-133604-0003__pitch+0__seed0__PML.wav", "WAV_sha256": "27b6d572359d76b35220df294a9cdb3e1e47e1b1edffde3f6d7151c8f2291e70", "bytes": 160044, "source_NPY_sha256": "d3e29b6ecedbb29f259d2b5da2790a9fb769b78cda50fe1daae75d47da0e9713", "condition_and_source_NPY_identity_matched": true, "excerpt_sample_range": [ 0, 80000 ] }, { "variant": "RAW_FULL", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/1188-133604-0003__pitch+0__seed0__RAW_FULL.wav", "WAV_sha256": "a0a1ee678817ac0e0b209a16d35c4c0602a317a33252de173fa59992cdaa98e6", "bytes": 160044, "source_NPY_sha256": "2b1046ba58dad230b4eb3350850b571682f9f862c5cb9f88b6ea1863ade21b3b", "condition_and_source_NPY_identity_matched": true, "excerpt_sample_range": [ 0, 80000 ] } ], "bytes": 160044, "WAV_sha256": "fddc422f733af73aae0daaba9b5e375a6c1bd03779153ddc168d5b4897a70a64", "header": { "container": "RIFF/WAVE", "format_code": 1, "format": "PCM", "endianness": "little", "channels": 1, "sample_rate": 16000, "byte_rate": 32000, "block_align": 2, "bits_per_sample": 16, "sample_width_bytes": 2, "samples": 80000, "header_bytes": 44, "riff_size": 160036, "data_bytes": 160000, "header_hex": "524946462471020057415645666d74201000000001000100803e0000007d0000020010006461746100710200", "header_sha256": "fe37d7895050fe71672e015eccb2ba6a4ca09fed7bdcd9a838adf2c905a2ca42", "PCM_data_sha256": "5bfe9585d941beff2ca2611e733663f346cdc6fae5bd0335de28fce88d68048d" }, "written_PCM_matches_declared_conversion": true }, { "case": "librispeech_121-121726-0012", "sample": 3, "recording_id": "121-121726-0012", "pitch_semitones": 0, "seed": 0, "variant": "RAW_POSTCOLOR_M1", "file": "audio/121-121726-0012__pitch+0__seed0__RAW_POSTCOLOR_M1.wav", "source_NPY_sha256": "6a130384d4ef170fe3271ba22740fa9fcf3f06937a9db63a274cd4a4d5157656", "source_NPY_file_sha256": "6a130384d4ef170fe3271ba22740fa9fcf3f06937a9db63a274cd4a4d5157656", "source_NPY_bytes": 517888, "source_array_sha256": "5409f2ef9834522ce75e107b7732b904be083891e4c2acf83a58b440ea41de5c", "source_array_sha256_definition": "NPYヘッダーを除く全長のlittle-endian float64標本列。C-orderの連続bytes。", "source_array_bytes": 517760, "source_dtype": "float64", "source_byte_order": "little", "source_samples": 64720, "source_sample_rate": 16000, "source_all_finite": true, "source_nonfinite_count": 0, "source_min": -0.5383137283547512, "source_max": 0.5782088684209157, "source_peak": 0.5782088684209157, "samples": 64720, "sample_rate": 16000, "channels": 1, "sample_width_bytes": 2, "excerpt_sample_range": [ 0, 64720 ], "excerpt_sample_range_end": "exclusive", "excerpt_seconds": [ 0, 4.045 ], "excerpt_peak": 0.5782088684209157, "cut_reason": "5秒以内のため元の標本列全体", "gain": 1, "normalization": "none", "dither": "none", "resampling": "none", "alignment": "none", "quantization": "rint(x*32767), nearest ties-to-even, little-endian signed PCM16", "clipping": "none", "clip_required": false, "clip_count": 0, "PCM_min": -17639, "PCM_max": 18946, "source_file_identity_matched": { "final_receipt": true, "all_30_waveforms_csv": true, "condition_measurement": true }, "baseline_comparisons": [ { "variant": "PML", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/121-121726-0012__pitch+0__seed0__PML.wav", "WAV_sha256": "0e7bc2ec7d16b6c31fc971104a7f09ed9c6134f6279fabb63241bb11541fd2ce", "bytes": 129484, "source_NPY_sha256": "bee6b53a7fe774160c03a93083e906a9b38cf640e006db02ed7ff8660b282ace", "condition_and_source_NPY_identity_matched": true, "excerpt_sample_range": [ 0, 64720 ] }, { "variant": "RAW_FULL", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/121-121726-0012__pitch+0__seed0__RAW_FULL.wav", "WAV_sha256": "be4ef209bb6cb8195b1b8a633cff5a7671f803e38526d011f4a54aba6335560f", "bytes": 129484, "source_NPY_sha256": "6cf3876756c3750b646ec255f2594c3fa5e5030e632c9bbeb1c8b015d9d660d4", "condition_and_source_NPY_identity_matched": true, "excerpt_sample_range": [ 0, 64720 ] } ], "bytes": 129484, "WAV_sha256": "22bc6b3e792aabdc57646a1980a728396a885849c2719cafcca97662b7b09db0", "header": { "container": "RIFF/WAVE", "format_code": 1, "format": "PCM", "endianness": "little", "channels": 1, "sample_rate": 16000, "byte_rate": 32000, "block_align": 2, "bits_per_sample": 16, "sample_width_bytes": 2, "samples": 64720, "header_bytes": 44, "riff_size": 129476, "data_bytes": 129440, "header_hex": "52494646c4f9010057415645666d74201000000001000100803e0000007d00000200100064617461a0f90100", "header_sha256": "1430049218770f1f842b7d92f13fc6a86ed5691e022e91f9ce5ddc8397e15f73", "PCM_data_sha256": "26e2901088848ba63eec23b0843c8cfac0a9c766df226d4a9f20960e18449aed" }, "written_PCM_matches_declared_conversion": true }, { "case": "librispeech_1221-135766-0005", "sample": 4, "recording_id": "1221-135766-0005", "pitch_semitones": 0, "seed": 0, "variant": "RAW_POSTCOLOR_M1", "file": "audio/1221-135766-0005__pitch+0__seed0__RAW_POSTCOLOR_M1.wav", "source_NPY_sha256": "2c0a14d656fa3567c9dc38368537769da88762f1d8bd1eb7759b0a9f41537735", "source_NPY_file_sha256": "2c0a14d656fa3567c9dc38368537769da88762f1d8bd1eb7759b0a9f41537735", "source_NPY_bytes": 768128, "source_array_sha256": "2ada60e2fb495b5afd2967431459348777c3504ca1200b9cdf5c98bfafcb594a", "source_array_sha256_definition": "NPYヘッダーを除く全長のlittle-endian float64標本列。C-orderの連続bytes。", "source_array_bytes": 768000, "source_dtype": "float64", "source_byte_order": "little", "source_samples": 96000, "source_sample_rate": 16000, "source_all_finite": true, "source_nonfinite_count": 0, "source_min": -0.3597473888694367, "source_max": 0.3167807006102521, "source_peak": 0.3597473888694367, "samples": 80000, "sample_rate": 16000, "channels": 1, "sample_width_bytes": 2, "excerpt_sample_range": [ 0, 80000 ], "excerpt_sample_range_end": "exclusive", "excerpt_seconds": [ 0, 5.0 ], "excerpt_peak": 0.3597473888694367, "cut_reason": "先頭5秒まで", "gain": 1, "normalization": "none", "dither": "none", "resampling": "none", "alignment": "none", "quantization": "rint(x*32767), nearest ties-to-even, little-endian signed PCM16", "clipping": "none", "clip_required": false, "clip_count": 0, "PCM_min": -11788, "PCM_max": 10380, "source_file_identity_matched": { "final_receipt": true, "all_30_waveforms_csv": true, "condition_measurement": true }, "baseline_comparisons": [ { "variant": "PML", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/1221-135766-0005__pitch+0__seed0__PML.wav", "WAV_sha256": "d31ad45a3ba55c856cf6010015a6b16c5bddea64b518309fef6b2db0ab2d0611", "bytes": 160044, "source_NPY_sha256": "82cfae761e5c0f68551f01def9e993c41545f7fc8d731b60e1305811d7f63067", "condition_and_source_NPY_identity_matched": true, "excerpt_sample_range": [ 0, 80000 ] }, { "variant": "RAW_FULL", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/1221-135766-0005__pitch+0__seed0__RAW_FULL.wav", "WAV_sha256": "c8d68c053cb8d282027b7de52953c8e2abf966d514b084763755213b8ca8482a", "bytes": 160044, "source_NPY_sha256": "b45a6380b5c8f17818af0dfae734ac4982a0ee6a4490f8dfa38759fe2f4e2cb5", "condition_and_source_NPY_identity_matched": true, "excerpt_sample_range": [ 0, 80000 ] } ], "bytes": 160044, "WAV_sha256": "6d05f6730cde91454334b7d4a62c30d559b950e9b559b217bb8f13a541d62005", "header": { "container": "RIFF/WAVE", "format_code": 1, "format": "PCM", "endianness": "little", "channels": 1, "sample_rate": 16000, "byte_rate": 32000, "block_align": 2, "bits_per_sample": 16, "sample_width_bytes": 2, "samples": 80000, "header_bytes": 44, "riff_size": 160036, "data_bytes": 160000, "header_hex": "524946462471020057415645666d74201000000001000100803e0000007d0000020010006461746100710200", "header_sha256": "fe37d7895050fe71672e015eccb2ba6a4ca09fed7bdcd9a838adf2c905a2ca42", "PCM_data_sha256": "38577ded8bb72923a1b5ce52df70cd402ceaed6a2941995bf5f47bc88cc5c807" }, "written_PCM_matches_declared_conversion": true }, { "case": "librispeech_1089-134686-0032", "sample": 5, "recording_id": "1089-134686-0032", "pitch_semitones": 0, "seed": 0, "variant": "RAW_POSTCOLOR_M1", "file": "audio/1089-134686-0032__pitch+0__seed0__RAW_POSTCOLOR_M1.wav", "source_NPY_sha256": "64142a0704ebcd7e46a990da2b7126f3d88884e793f5287548fa9fe90b6c4d45", "source_NPY_file_sha256": "64142a0704ebcd7e46a990da2b7126f3d88884e793f5287548fa9fe90b6c4d45", "source_NPY_bytes": 523648, "source_array_sha256": "2c81bf21acd95e2096dab7e2539eb04de79cd6cabb1609da97c58df1a7013a05", "source_array_sha256_definition": "NPYヘッダーを除く全長のlittle-endian float64標本列。C-orderの連続bytes。", "source_array_bytes": 523520, "source_dtype": "float64", "source_byte_order": "little", "source_samples": 65440, "source_sample_rate": 16000, "source_all_finite": true, "source_nonfinite_count": 0, "source_min": -0.5224749812678056, "source_max": 0.33737706393003464, "source_peak": 0.5224749812678056, "samples": 65440, "sample_rate": 16000, "channels": 1, "sample_width_bytes": 2, "excerpt_sample_range": [ 0, 65440 ], "excerpt_sample_range_end": "exclusive", "excerpt_seconds": [ 0, 4.09 ], "excerpt_peak": 0.5224749812678056, "cut_reason": "5秒以内のため元の標本列全体", "gain": 1, "normalization": "none", "dither": "none", "resampling": "none", "alignment": "none", "quantization": "rint(x*32767), nearest ties-to-even, little-endian signed PCM16", "clipping": "none", "clip_required": false, "clip_count": 0, "PCM_min": -17120, "PCM_max": 11055, "source_file_identity_matched": { "final_receipt": true, "all_30_waveforms_csv": true, "condition_measurement": true }, "baseline_comparisons": [ { "variant": "PML", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/1089-134686-0032__pitch+0__seed0__PML.wav", "WAV_sha256": "c06b08e00da7f17b98a09810fec60b56da1e9d988a3e9ef7f7e8515965a82e1c", "bytes": 130924, "source_NPY_sha256": "80bf274d7883bff30b22f1eae1ea2a40cb7d3f7fdc0bc75162dc90e44faa6cc8", "condition_and_source_NPY_identity_matched": true, "excerpt_sample_range": [ 0, 65440 ] }, { "variant": "RAW_FULL", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/1089-134686-0032__pitch+0__seed0__RAW_FULL.wav", "WAV_sha256": "857563fdcff718426f63a7998c22ac42001835ed490b7ce35d245be6b5135fdf", "bytes": 130924, "source_NPY_sha256": "949f626d7910d668b396b77782f8be4f4609bfac31d94289026258b8bb674406", "condition_and_source_NPY_identity_matched": true, "excerpt_sample_range": [ 0, 65440 ] } ], "bytes": 130924, "WAV_sha256": "fc0bad1ea4e3e20c3440dec3b9d2f0941a6d25d5879a7628c531ac2d26099f2f", "header": { "container": "RIFF/WAVE", "format_code": 1, "format": "PCM", "endianness": "little", "channels": 1, "sample_rate": 16000, "byte_rate": 32000, "block_align": 2, "bits_per_sample": 16, "sample_width_bytes": 2, "samples": 65440, "header_bytes": 44, "riff_size": 130916, "data_bytes": 130880, "header_hex": "5249464664ff010057415645666d74201000000001000100803e0000007d0000020010006461746140ff0100", "header_sha256": "40f6385e2c12909fb203eff75e76d10044107b36449e5cbae62b3f488a6b7bbe", "PCM_data_sha256": "f93fd7009ad5faae57e2ea2e4425357fe22947f03fbc05eeb469be59dbbdca7d" }, "written_PCM_matches_declared_conversion": true }, { "case": "librispeech_121-121726-0012", "sample": 3, "recording_id": "121-121726-0012", "pitch_semitones": -3, "seed": 0, "variant": "RAW_POSTCOLOR_M1", "file": "audio/121-121726-0012__pitch-3__seed0__RAW_POSTCOLOR_M1.wav", "source_NPY_sha256": "6d4606089c4c9428cebe258d7036c74c04e13b155312e7bc4e5c03fa4559ca72", "source_NPY_file_sha256": "6d4606089c4c9428cebe258d7036c74c04e13b155312e7bc4e5c03fa4559ca72", "source_NPY_bytes": 517888, "source_array_sha256": "2fccb1ba7017b5ede5e41b955c3b23ea5f8706042f8305bb43df088656f45ea7", "source_array_sha256_definition": "NPYヘッダーを除く全長のlittle-endian float64標本列。C-orderの連続bytes。", "source_array_bytes": 517760, "source_dtype": "float64", "source_byte_order": "little", "source_samples": 64720, "source_sample_rate": 16000, "source_all_finite": true, "source_nonfinite_count": 0, "source_min": -0.4868027118932332, "source_max": 0.6000559372364478, "source_peak": 0.6000559372364478, "samples": 64720, "sample_rate": 16000, "channels": 1, "sample_width_bytes": 2, "excerpt_sample_range": [ 0, 64720 ], "excerpt_sample_range_end": "exclusive", "excerpt_seconds": [ 0, 4.045 ], "excerpt_peak": 0.6000559372364478, "cut_reason": "5秒以内のため元の標本列全体", "gain": 1, "normalization": "none", "dither": "none", "resampling": "none", "alignment": "none", "quantization": "rint(x*32767), nearest ties-to-even, little-endian signed PCM16", "clipping": "none", "clip_required": false, "clip_count": 0, "PCM_min": -15951, "PCM_max": 19662, "source_file_identity_matched": { "final_receipt": true, "all_30_waveforms_csv": true, "condition_measurement": true }, "baseline_comparisons": [ { "variant": "PML", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/121-121726-0012__pitch-3__seed0__PML.wav", "WAV_sha256": "18e7dd0140ad704e8dc15069a7e7466d77983d1a8e0bd32aecdde1a983e5b53f", "bytes": 129484, "source_NPY_sha256": "b5d58d1a3452c7cced18f708f3b23c109acb016e90f0ab159606d447fa2a8ba2", "condition_and_source_NPY_identity_matched": true, "excerpt_sample_range": [ 0, 64720 ] }, { "variant": "RAW_FULL", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/121-121726-0012__pitch-3__seed0__RAW_FULL.wav", "WAV_sha256": "8d8f880ea646a1a25e7555afdf936d47ac4371be2e538c12cb41de2a33540666", "bytes": 129484, "source_NPY_sha256": "c8da7838eee5e846ea1ea3e638a4f7270f7be433fc9c893642421fd34bbe512e", "condition_and_source_NPY_identity_matched": true, "excerpt_sample_range": [ 0, 64720 ] } ], "bytes": 129484, "WAV_sha256": "bb6dfa009caffffba2968e9994769f0b90801fd1ff3d66024cf9859242140c91", "header": { "container": "RIFF/WAVE", "format_code": 1, "format": "PCM", "endianness": "little", "channels": 1, "sample_rate": 16000, "byte_rate": 32000, "block_align": 2, "bits_per_sample": 16, "sample_width_bytes": 2, "samples": 64720, "header_bytes": 44, "riff_size": 129476, "data_bytes": 129440, "header_hex": "52494646c4f9010057415645666d74201000000001000100803e0000007d00000200100064617461a0f90100", "header_sha256": "1430049218770f1f842b7d92f13fc6a86ed5691e022e91f9ce5ddc8397e15f73", "PCM_data_sha256": "b90f5544a2e703444836297fdedce694b4757703250782c06791d48ecf075674" }, "written_PCM_matches_declared_conversion": true }, { "case": "librispeech_121-121726-0012", "sample": 3, "recording_id": "121-121726-0012", "pitch_semitones": -3, "seed": 20261015, "variant": "RAW_POSTCOLOR_M1", "file": "audio/121-121726-0012__pitch-3__seed20261015__RAW_POSTCOLOR_M1.wav", "source_NPY_sha256": "e97ff7851362c2039165cb37955753f1497eb355c8fa85b620deef3e8bb4094b", "source_NPY_file_sha256": "e97ff7851362c2039165cb37955753f1497eb355c8fa85b620deef3e8bb4094b", "source_NPY_bytes": 517888, "source_array_sha256": "4d93ad0c75706dae8248e12cca4a0a16e795ef1c4441b046c5d4159de238bc01", "source_array_sha256_definition": "NPYヘッダーを除く全長のlittle-endian float64標本列。C-orderの連続bytes。", "source_array_bytes": 517760, "source_dtype": "float64", "source_byte_order": "little", "source_samples": 64720, "source_sample_rate": 16000, "source_all_finite": true, "source_nonfinite_count": 0, "source_min": -0.4814744289033115, "source_max": 0.6727290041449123, "source_peak": 0.6727290041449123, "samples": 64720, "sample_rate": 16000, "channels": 1, "sample_width_bytes": 2, "excerpt_sample_range": [ 0, 64720 ], "excerpt_sample_range_end": "exclusive", "excerpt_seconds": [ 0, 4.045 ], "excerpt_peak": 0.6727290041449123, "cut_reason": "5秒以内のため元の標本列全体", "gain": 1, "normalization": "none", "dither": "none", "resampling": "none", "alignment": "none", "quantization": "rint(x*32767), nearest ties-to-even, little-endian signed PCM16", "clipping": "none", "clip_required": false, "clip_count": 0, "PCM_min": -15776, "PCM_max": 22043, "source_file_identity_matched": { "final_receipt": true, "all_30_waveforms_csv": true, "condition_measurement": true }, "baseline_comparisons": [ { "variant": "PML", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/121-121726-0012__pitch-3__seed20261015__PML.wav", "WAV_sha256": "169de85768a72963c76c17815f3ab61a18e40b1c3057e5a63ef39e972795fd71", "bytes": 129484, "source_NPY_sha256": "2ed85deb832fe249f779569e7c6379b57d3db0525df04a26444f4d5151db0b89", "condition_and_source_NPY_identity_matched": true, "excerpt_sample_range": [ 0, 64720 ] }, { "variant": "RAW_FULL", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/121-121726-0012__pitch-3__seed20261015__RAW_FULL.wav", "WAV_sha256": "617484725c170f40b86cefca5a42ab00ae91a455916e31cc4bd4b31f06ac241c", "bytes": 129484, "source_NPY_sha256": "5301869c730597ada086b2a0b7274654e22a93872364e3a7868e1d74098902c2", "condition_and_source_NPY_identity_matched": true, "excerpt_sample_range": [ 0, 64720 ] } ], "bytes": 129484, "WAV_sha256": "d5e1c38c17eb684078da2578754b71a226f8ebccf3eb5d81a92bed2961f6f329", "header": { "container": "RIFF/WAVE", "format_code": 1, "format": "PCM", "endianness": "little", "channels": 1, "sample_rate": 16000, "byte_rate": 32000, "block_align": 2, "bits_per_sample": 16, "sample_width_bytes": 2, "samples": 64720, "header_bytes": 44, "riff_size": 129476, "data_bytes": 129440, "header_hex": "52494646c4f9010057415645666d74201000000001000100803e0000007d00000200100064617461a0f90100", "header_sha256": "1430049218770f1f842b7d92f13fc6a86ed5691e022e91f9ce5ddc8397e15f73", "PCM_data_sha256": "0f77b112656c406be8d1563e8411374cdff8080191e26bac6ce1261609336c78" }, "written_PCM_matches_declared_conversion": true }, { "case": "librispeech_121-121726-0012", "sample": 3, "recording_id": "121-121726-0012", "pitch_semitones": 3, "seed": 0, "variant": "RAW_POSTCOLOR_M1", "file": "audio/121-121726-0012__pitch+3__seed0__RAW_POSTCOLOR_M1.wav", "source_NPY_sha256": "46052d51861fb74a1be95700ad4031a521d1a9ce5fe76c268587762098b83f69", "source_NPY_file_sha256": "46052d51861fb74a1be95700ad4031a521d1a9ce5fe76c268587762098b83f69", "source_NPY_bytes": 517888, "source_array_sha256": "526caea6be0a521c51ef1ae68f5d9b77e509e9343b0125ead3dd8d848e337a7f", "source_array_sha256_definition": "NPYヘッダーを除く全長のlittle-endian float64標本列。C-orderの連続bytes。", "source_array_bytes": 517760, "source_dtype": "float64", "source_byte_order": "little", "source_samples": 64720, "source_sample_rate": 16000, "source_all_finite": true, "source_nonfinite_count": 0, "source_min": -0.5097642851426372, "source_max": 0.49401614668101956, "source_peak": 0.5097642851426372, "samples": 64720, "sample_rate": 16000, "channels": 1, "sample_width_bytes": 2, "excerpt_sample_range": [ 0, 64720 ], "excerpt_sample_range_end": "exclusive", "excerpt_seconds": [ 0, 4.045 ], "excerpt_peak": 0.5097642851426372, "cut_reason": "5秒以内のため元の標本列全体", "gain": 1, "normalization": "none", "dither": "none", "resampling": "none", "alignment": "none", "quantization": "rint(x*32767), nearest ties-to-even, little-endian signed PCM16", "clipping": "none", "clip_required": false, "clip_count": 0, "PCM_min": -16703, "PCM_max": 16187, "source_file_identity_matched": { "final_receipt": true, "all_30_waveforms_csv": true, "condition_measurement": true }, "baseline_comparisons": [ { "variant": "PML", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/121-121726-0012__pitch+3__seed0__PML.wav", "WAV_sha256": "d1fcaee4b6a8981f09e6b6b8583ad741e7ceaeb766e1a8b913b64f13dc88746a", "bytes": 129484, "source_NPY_sha256": "082bd02b2d99751a844f503d91a3c6df5c7f98d60aef01c2c8d9ee481295c776", "condition_and_source_NPY_identity_matched": true, "excerpt_sample_range": [ 0, 64720 ] }, { "variant": "RAW_FULL", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/121-121726-0012__pitch+3__seed0__RAW_FULL.wav", "WAV_sha256": "8b97a060d79d1f74bee0a6cd3f3e267d400be0c988b5890eba436bddcc21ccee", "bytes": 129484, "source_NPY_sha256": "448a41231e539e37d40953699caf50f78ed0866197fa9c0aa01c689ca773f9f8", "condition_and_source_NPY_identity_matched": true, "excerpt_sample_range": [ 0, 64720 ] } ], "bytes": 129484, "WAV_sha256": "ab394a1783b9f9e38a22c4f913c47c7dc46ad67efa959f8ab1367b773439a613", "header": { "container": "RIFF/WAVE", "format_code": 1, "format": "PCM", "endianness": "little", "channels": 1, "sample_rate": 16000, "byte_rate": 32000, "block_align": 2, "bits_per_sample": 16, "sample_width_bytes": 2, "samples": 64720, "header_bytes": 44, "riff_size": 129476, "data_bytes": 129440, "header_hex": "52494646c4f9010057415645666d74201000000001000100803e0000007d00000200100064617461a0f90100", "header_sha256": "1430049218770f1f842b7d92f13fc6a86ed5691e022e91f9ce5ddc8397e15f73", "PCM_data_sha256": "66019b6b0415ac2bfd431930146a0478d9b47a01717222108f56355bf59c5e9d" }, "written_PCM_matches_declared_conversion": true }, { "case": "librispeech_121-121726-0012", "sample": 3, "recording_id": "121-121726-0012", "pitch_semitones": 3, "seed": 20261015, "variant": "RAW_POSTCOLOR_M1", "file": "audio/121-121726-0012__pitch+3__seed20261015__RAW_POSTCOLOR_M1.wav", "source_NPY_sha256": "934430303be39d2600572c3b27ec5a10c7da2e3e02492fff9eb6fb328da612ee", "source_NPY_file_sha256": "934430303be39d2600572c3b27ec5a10c7da2e3e02492fff9eb6fb328da612ee", "source_NPY_bytes": 517888, "source_array_sha256": "2004b9eb72fa5b7d438b65fd4287d8d696ce52688ecdcb6ecb187acba0e77fae", "source_array_sha256_definition": "NPYヘッダーを除く全長のlittle-endian float64標本列。C-orderの連続bytes。", "source_array_bytes": 517760, "source_dtype": "float64", "source_byte_order": "little", "source_samples": 64720, "source_sample_rate": 16000, "source_all_finite": true, "source_nonfinite_count": 0, "source_min": -0.4608520793021853, "source_max": 0.8226684713008581, "source_peak": 0.8226684713008581, "samples": 64720, "sample_rate": 16000, "channels": 1, "sample_width_bytes": 2, "excerpt_sample_range": [ 0, 64720 ], "excerpt_sample_range_end": "exclusive", "excerpt_seconds": [ 0, 4.045 ], "excerpt_peak": 0.8226684713008581, "cut_reason": "5秒以内のため元の標本列全体", "gain": 1, "normalization": "none", "dither": "none", "resampling": "none", "alignment": "none", "quantization": "rint(x*32767), nearest ties-to-even, little-endian signed PCM16", "clipping": "none", "clip_required": false, "clip_count": 0, "PCM_min": -15101, "PCM_max": 26956, "source_file_identity_matched": { "final_receipt": true, "all_30_waveforms_csv": true, "condition_measurement": true }, "baseline_comparisons": [ { "variant": "PML", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/121-121726-0012__pitch+3__seed20261015__PML.wav", "WAV_sha256": "1e815a8c64819cc373635ee742c8c0debaa1934fddd47b183d3769ccff0e61f3", "bytes": 129484, "source_NPY_sha256": "d56d4907bb8ee291d7ca7ff2c240ef4a1bdb18d485c78449b45dd152818c0e26", "condition_and_source_NPY_identity_matched": true, "excerpt_sample_range": [ 0, 64720 ] }, { "variant": "RAW_FULL", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/121-121726-0012__pitch+3__seed20261015__RAW_FULL.wav", "WAV_sha256": "95a05df2003897deac9c19b0cbc778bf928136f828748add8113707d6a8ec2c9", "bytes": 129484, "source_NPY_sha256": "53a9de0adae6d1f1b59307c0e9d5345f9cbdd7fc155be3ba20e9327d5ae7085b", "condition_and_source_NPY_identity_matched": true, "excerpt_sample_range": [ 0, 64720 ] } ], "bytes": 129484, "WAV_sha256": "dbea73251a46a295264687dabaf5e0531c86399a104aeb93689b9f429b734b0a", "header": { "container": "RIFF/WAVE", "format_code": 1, "format": "PCM", "endianness": "little", "channels": 1, "sample_rate": 16000, "byte_rate": 32000, "block_align": 2, "bits_per_sample": 16, "sample_width_bytes": 2, "samples": 64720, "header_bytes": 44, "riff_size": 129476, "data_bytes": 129440, "header_hex": "52494646c4f9010057415645666d74201000000001000100803e0000007d00000200100064617461a0f90100", "header_sha256": "1430049218770f1f842b7d92f13fc6a86ed5691e022e91f9ce5ddc8397e15f73", "PCM_data_sha256": "3fc0c9d59d66adff978ec6e61e48e398db730d68909e1565301493a88ac6f1e8" }, "written_PCM_matches_declared_conversion": true } ], "existing_public_references": [ { "case": "librispeech_1089-134686-0008", "pitch_semitones": null, "seed": null, "variant": "ORIGINAL", "bytes": 160044, "WAV_sha256": "2f867a9dafd1ca92b56d0dde8b48be73ebeb328fc03ef3ce032fd5be18963af0", "source_NPY_sha256": "2906c24c57cf16b7a1a29dac15e0d727cb8b4679724efec0f565946723cc9fd8", "samples": 80000, "sample_rate": 16000, "channels": 1, "source_WAV_sha256": "cf2033d5f036076f2dc9613fd5c90f7a2bf4cc31293bf2e680a201914ad4a5f1", "source_samples": 96000, "excerpt_sample_range": [ 0, 80000 ], "excerpt_seconds": [ 0, 5.0 ], "PCM_samples_unchanged": true, "cut_reason": "Publication route maximum five seconds", "sample_width_bytes": 2, "gain": 1, "normalization": "none", "clipping": "none", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/1089-134686-0008__original.wav", "copied_to_this_report": false, "header": { "container": "RIFF/WAVE", "format_code": 1, "format": "PCM", "endianness": "little", "channels": 1, "sample_rate": 16000, "byte_rate": 32000, "block_align": 2, "bits_per_sample": 16, "sample_width_bytes": 2, "samples": 80000, "header_bytes": 44, "riff_size": 160036, "data_bytes": 160000, "header_hex": "524946462471020057415645666d74201000000001000100803e0000007d0000020010006461746100710200", "header_sha256": "fe37d7895050fe71672e015eccb2ba6a4ca09fed7bdcd9a838adf2c905a2ca42", "PCM_data_sha256": "4125b70d2329cee2b0815d993f5dd9d2be0d6aff3179d5a53119408e58e7588e" }, "reference_role": "LibriSpeech原音は内容・時刻の参考。音高編集音の正解音声ではない。" }, { "case": "librispeech_1089-134686-0008", "pitch_semitones": 0, "seed": 0, "variant": "PML", "bytes": 160044, "WAV_sha256": "54d2b141b5b7e67bc735ca5e399f54a935853956e8f5ff7286b1651865f71d51", "source_NPY_sha256": "66fa9fe6bb6e0880a2213e700ec59f6dd8af48283c5c350d980d948bf12644f6", "samples": 80000, "sample_rate": 16000, "channels": 1, "source_WAV_sha256": "4e72015f296dffb979d3f2bcb84d64b9b75ad0016bdd9da37f32f0923e418509", "source_samples": 96000, "excerpt_sample_range": [ 0, 80000 ], "excerpt_seconds": [ 0, 5.0 ], "PCM_samples_unchanged": true, "cut_reason": "Publication route maximum five seconds", "sample_width_bytes": 2, "gain": 1, "normalization": "none", "clipping": "none", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/1089-134686-0008__pitch+0__seed0__PML.wav", "copied_to_this_report": false, "header": { "container": "RIFF/WAVE", "format_code": 1, "format": "PCM", "endianness": "little", "channels": 1, "sample_rate": 16000, "byte_rate": 32000, "block_align": 2, "bits_per_sample": 16, "sample_width_bytes": 2, "samples": 80000, "header_bytes": 44, "riff_size": 160036, "data_bytes": 160000, "header_hex": "524946462471020057415645666d74201000000001000100803e0000007d0000020010006461746100710200", "header_sha256": "fe37d7895050fe71672e015eccb2ba6a4ca09fed7bdcd9a838adf2c905a2ca42", "PCM_data_sha256": "6b5cba97dc0b1a2827fe61e87ad9ba582b3ededa5d48673a625eda0d8e84a5cd" } }, { "case": "librispeech_1089-134686-0008", "pitch_semitones": 0, "seed": 0, "variant": "RAW_FULL", "bytes": 160044, "WAV_sha256": "eb7822699fd9ed0aeb2382a2edb2a3a358541787286b74f46eaceb01add4b9e1", "source_NPY_sha256": "64a0d355df20f0bc11c6826f22412a88cb5259f79a7f1b115ad836c18fe43376", "samples": 80000, "sample_rate": 16000, "channels": 1, "source_WAV_sha256": "e89bd22fe738ee8692e44b435458a1e9219d5daa6e0ce17cbc733cd7e47aae93", "source_samples": 96000, "excerpt_sample_range": [ 0, 80000 ], "excerpt_seconds": [ 0, 5.0 ], "PCM_samples_unchanged": true, "cut_reason": "Publication route maximum five seconds", "sample_width_bytes": 2, "gain": 1, "normalization": "none", "clipping": "none", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/1089-134686-0008__pitch+0__seed0__RAW_FULL.wav", "copied_to_this_report": false, "header": { "container": "RIFF/WAVE", "format_code": 1, "format": "PCM", "endianness": "little", "channels": 1, "sample_rate": 16000, "byte_rate": 32000, "block_align": 2, "bits_per_sample": 16, "sample_width_bytes": 2, "samples": 80000, "header_bytes": 44, "riff_size": 160036, "data_bytes": 160000, "header_hex": "524946462471020057415645666d74201000000001000100803e0000007d0000020010006461746100710200", "header_sha256": "fe37d7895050fe71672e015eccb2ba6a4ca09fed7bdcd9a838adf2c905a2ca42", "PCM_data_sha256": "d335269fe13088519e7b874b70dd780ac31b3de775112b666c1f73461ca6fb88" } }, { "case": "librispeech_1188-133604-0003", "pitch_semitones": null, "seed": null, "variant": "ORIGINAL", "bytes": 160044, "WAV_sha256": "e2273742018cd38e5af7b214d8d36fd3f9daf72709a288e49133d3622f442c90", "source_NPY_sha256": "d03996160873a22360bde5b073fc057837439fa39b35136ceaeb8ffda73d15c2", "samples": 80000, "sample_rate": 16000, "channels": 1, "source_WAV_sha256": "84d54b72ca45d33ed359712eea5d49739535969db269156b05a159b1f1ab5aeb", "source_samples": 96000, "excerpt_sample_range": [ 0, 80000 ], "excerpt_seconds": [ 0, 5.0 ], "PCM_samples_unchanged": true, "cut_reason": "Publication route maximum five seconds", "sample_width_bytes": 2, "gain": 1, "normalization": "none", "clipping": "none", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/1188-133604-0003__original.wav", "copied_to_this_report": false, "header": { "container": "RIFF/WAVE", "format_code": 1, "format": "PCM", "endianness": "little", "channels": 1, "sample_rate": 16000, "byte_rate": 32000, "block_align": 2, "bits_per_sample": 16, "sample_width_bytes": 2, "samples": 80000, "header_bytes": 44, "riff_size": 160036, "data_bytes": 160000, "header_hex": "524946462471020057415645666d74201000000001000100803e0000007d0000020010006461746100710200", "header_sha256": "fe37d7895050fe71672e015eccb2ba6a4ca09fed7bdcd9a838adf2c905a2ca42", "PCM_data_sha256": "cf8c0c3e8ecdd10966dd869d92ec99ad42e6db8821b5fdc702d2084d3188e039" }, "reference_role": "LibriSpeech原音は内容・時刻の参考。音高編集音の正解音声ではない。" }, { "case": "librispeech_1188-133604-0003", "pitch_semitones": 0, "seed": 0, "variant": "PML", "bytes": 160044, "WAV_sha256": "27b6d572359d76b35220df294a9cdb3e1e47e1b1edffde3f6d7151c8f2291e70", "source_NPY_sha256": "d3e29b6ecedbb29f259d2b5da2790a9fb769b78cda50fe1daae75d47da0e9713", "samples": 80000, "sample_rate": 16000, "channels": 1, "source_WAV_sha256": "aa424f511eec2206563d4b6d6afbde1e78e1faa74478d7a53d470bbe280e0e01", "source_samples": 96000, "excerpt_sample_range": [ 0, 80000 ], "excerpt_seconds": [ 0, 5.0 ], "PCM_samples_unchanged": true, "cut_reason": "Publication route maximum five seconds", "sample_width_bytes": 2, "gain": 1, "normalization": "none", "clipping": "none", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/1188-133604-0003__pitch+0__seed0__PML.wav", "copied_to_this_report": false, "header": { "container": "RIFF/WAVE", "format_code": 1, "format": "PCM", "endianness": "little", "channels": 1, "sample_rate": 16000, "byte_rate": 32000, "block_align": 2, "bits_per_sample": 16, "sample_width_bytes": 2, "samples": 80000, "header_bytes": 44, "riff_size": 160036, "data_bytes": 160000, "header_hex": "524946462471020057415645666d74201000000001000100803e0000007d0000020010006461746100710200", "header_sha256": "fe37d7895050fe71672e015eccb2ba6a4ca09fed7bdcd9a838adf2c905a2ca42", "PCM_data_sha256": "bf5d8d2fcb0c3724cde97b0330c0a5473ee2043796504ab0e96f7db8c7c8ad91" } }, { "case": "librispeech_1188-133604-0003", "pitch_semitones": 0, "seed": 0, "variant": "RAW_FULL", "bytes": 160044, "WAV_sha256": "a0a1ee678817ac0e0b209a16d35c4c0602a317a33252de173fa59992cdaa98e6", "source_NPY_sha256": "2b1046ba58dad230b4eb3350850b571682f9f862c5cb9f88b6ea1863ade21b3b", "samples": 80000, "sample_rate": 16000, "channels": 1, "source_WAV_sha256": "21a483309a82217c941da6bba1289b54601014c04d814f9e16a2711d93a773b4", "source_samples": 96000, "excerpt_sample_range": [ 0, 80000 ], "excerpt_seconds": [ 0, 5.0 ], "PCM_samples_unchanged": true, "cut_reason": "Publication route maximum five seconds", "sample_width_bytes": 2, "gain": 1, "normalization": "none", "clipping": "none", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/1188-133604-0003__pitch+0__seed0__RAW_FULL.wav", "copied_to_this_report": false, "header": { "container": "RIFF/WAVE", "format_code": 1, "format": "PCM", "endianness": "little", "channels": 1, "sample_rate": 16000, "byte_rate": 32000, "block_align": 2, "bits_per_sample": 16, "sample_width_bytes": 2, "samples": 80000, "header_bytes": 44, "riff_size": 160036, "data_bytes": 160000, "header_hex": "524946462471020057415645666d74201000000001000100803e0000007d0000020010006461746100710200", "header_sha256": "fe37d7895050fe71672e015eccb2ba6a4ca09fed7bdcd9a838adf2c905a2ca42", "PCM_data_sha256": "25ad11fc17c16155647823e962fa29206f8fec857abf73a23d3d51811246ee30" } }, { "case": "librispeech_121-121726-0012", "pitch_semitones": null, "seed": null, "variant": "ORIGINAL", "bytes": 129484, "WAV_sha256": "f509e04154bfad1df61d2b56feaa41a3b1474f35d986aa225d05a5b9e9cef1b6", "source_NPY_sha256": "d7b6a60f190aeff3e64279914fcc2517ac55145c4993566d15cb7fa6c01169ef", "samples": 64720, "sample_rate": 16000, "channels": 1, "source_WAV_sha256": "f509e04154bfad1df61d2b56feaa41a3b1474f35d986aa225d05a5b9e9cef1b6", "source_samples": 64720, "excerpt_sample_range": [ 0, 64720 ], "excerpt_seconds": [ 0, 4.045 ], "PCM_samples_unchanged": true, "cut_reason": "Full waveform within five seconds", "sample_width_bytes": 2, "gain": 1, "normalization": "none", "clipping": "none", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/121-121726-0012__original.wav", "copied_to_this_report": false, "header": { "container": "RIFF/WAVE", "format_code": 1, "format": "PCM", "endianness": "little", "channels": 1, "sample_rate": 16000, "byte_rate": 32000, "block_align": 2, "bits_per_sample": 16, "sample_width_bytes": 2, "samples": 64720, "header_bytes": 44, "riff_size": 129476, "data_bytes": 129440, "header_hex": "52494646c4f9010057415645666d74201000000001000100803e0000007d00000200100064617461a0f90100", "header_sha256": "1430049218770f1f842b7d92f13fc6a86ed5691e022e91f9ce5ddc8397e15f73", "PCM_data_sha256": "d3deb6e885601c0774b572f6302eff729a590aa24ab8c87d1f5e3e34575aa081" }, "reference_role": "LibriSpeech原音は内容・時刻の参考。音高編集音の正解音声ではない。" }, { "case": "librispeech_121-121726-0012", "pitch_semitones": 0, "seed": 0, "variant": "PML", "bytes": 129484, "WAV_sha256": "0e7bc2ec7d16b6c31fc971104a7f09ed9c6134f6279fabb63241bb11541fd2ce", "source_NPY_sha256": "bee6b53a7fe774160c03a93083e906a9b38cf640e006db02ed7ff8660b282ace", "samples": 64720, "sample_rate": 16000, "channels": 1, "source_WAV_sha256": "0e7bc2ec7d16b6c31fc971104a7f09ed9c6134f6279fabb63241bb11541fd2ce", "source_samples": 64720, "excerpt_sample_range": [ 0, 64720 ], "excerpt_seconds": [ 0, 4.045 ], "PCM_samples_unchanged": true, "cut_reason": "Full waveform within five seconds", "sample_width_bytes": 2, "gain": 1, "normalization": "none", "clipping": "none", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/121-121726-0012__pitch+0__seed0__PML.wav", "copied_to_this_report": false, "header": { "container": "RIFF/WAVE", "format_code": 1, "format": "PCM", "endianness": "little", "channels": 1, "sample_rate": 16000, "byte_rate": 32000, "block_align": 2, "bits_per_sample": 16, "sample_width_bytes": 2, "samples": 64720, "header_bytes": 44, "riff_size": 129476, "data_bytes": 129440, "header_hex": "52494646c4f9010057415645666d74201000000001000100803e0000007d00000200100064617461a0f90100", "header_sha256": "1430049218770f1f842b7d92f13fc6a86ed5691e022e91f9ce5ddc8397e15f73", "PCM_data_sha256": "aeb708bed6b170a7ff1909791132fbff2ba20b99265ceaa231952b774d92346b" } }, { "case": "librispeech_121-121726-0012", "pitch_semitones": 0, "seed": 0, "variant": "RAW_FULL", "bytes": 129484, "WAV_sha256": "be4ef209bb6cb8195b1b8a633cff5a7671f803e38526d011f4a54aba6335560f", "source_NPY_sha256": "6cf3876756c3750b646ec255f2594c3fa5e5030e632c9bbeb1c8b015d9d660d4", "samples": 64720, "sample_rate": 16000, "channels": 1, "source_WAV_sha256": "be4ef209bb6cb8195b1b8a633cff5a7671f803e38526d011f4a54aba6335560f", "source_samples": 64720, "excerpt_sample_range": [ 0, 64720 ], "excerpt_seconds": [ 0, 4.045 ], "PCM_samples_unchanged": true, "cut_reason": "Full waveform within five seconds", "sample_width_bytes": 2, "gain": 1, "normalization": "none", "clipping": "none", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/121-121726-0012__pitch+0__seed0__RAW_FULL.wav", "copied_to_this_report": false, "header": { "container": "RIFF/WAVE", "format_code": 1, "format": "PCM", "endianness": "little", "channels": 1, "sample_rate": 16000, "byte_rate": 32000, "block_align": 2, "bits_per_sample": 16, "sample_width_bytes": 2, "samples": 64720, "header_bytes": 44, "riff_size": 129476, "data_bytes": 129440, "header_hex": "52494646c4f9010057415645666d74201000000001000100803e0000007d00000200100064617461a0f90100", "header_sha256": "1430049218770f1f842b7d92f13fc6a86ed5691e022e91f9ce5ddc8397e15f73", "PCM_data_sha256": "9b9f545301da1d667269fc4a771632f7e81203aff488cc3bcb694562f485c187" } }, { "case": "librispeech_1221-135766-0005", "pitch_semitones": null, "seed": null, "variant": "ORIGINAL", "bytes": 160044, "WAV_sha256": "e3838c3f536df63fff50e941dfa3422e95acc8b1c70716779dbbb4d7b6733cb6", "source_NPY_sha256": "bc7fa6d71f6aee37b8d76a9a811d92b77e57fed28d8e0624cf7dff4190039e39", "samples": 80000, "sample_rate": 16000, "channels": 1, "source_WAV_sha256": "2be7bfdf2803931e3aa5cb7d0a175e9d2926cca0701b42ffafefa30ebc1f5d22", "source_samples": 96000, "excerpt_sample_range": [ 0, 80000 ], "excerpt_seconds": [ 0, 5.0 ], "PCM_samples_unchanged": true, "cut_reason": "Publication route maximum five seconds", "sample_width_bytes": 2, "gain": 1, "normalization": "none", "clipping": "none", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/1221-135766-0005__original.wav", "copied_to_this_report": false, "header": { "container": "RIFF/WAVE", "format_code": 1, "format": "PCM", "endianness": "little", "channels": 1, "sample_rate": 16000, "byte_rate": 32000, "block_align": 2, "bits_per_sample": 16, "sample_width_bytes": 2, "samples": 80000, "header_bytes": 44, "riff_size": 160036, "data_bytes": 160000, "header_hex": "524946462471020057415645666d74201000000001000100803e0000007d0000020010006461746100710200", "header_sha256": "fe37d7895050fe71672e015eccb2ba6a4ca09fed7bdcd9a838adf2c905a2ca42", "PCM_data_sha256": "926cccee29b08be01e950fcb6f9556a6826dc739a0cfcfcc46372a9af38c1b15" }, "reference_role": "LibriSpeech原音は内容・時刻の参考。音高編集音の正解音声ではない。" }, { "case": "librispeech_1221-135766-0005", "pitch_semitones": 0, "seed": 0, "variant": "PML", "bytes": 160044, "WAV_sha256": "d31ad45a3ba55c856cf6010015a6b16c5bddea64b518309fef6b2db0ab2d0611", "source_NPY_sha256": "82cfae761e5c0f68551f01def9e993c41545f7fc8d731b60e1305811d7f63067", "samples": 80000, "sample_rate": 16000, "channels": 1, "source_WAV_sha256": "d7e4aff5eb093ef42536fdf9fc220c5dd6a146e72cdf7a4683325dbb7987602e", "source_samples": 96000, "excerpt_sample_range": [ 0, 80000 ], "excerpt_seconds": [ 0, 5.0 ], "PCM_samples_unchanged": true, "cut_reason": "Publication route maximum five seconds", "sample_width_bytes": 2, "gain": 1, "normalization": "none", "clipping": "none", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/1221-135766-0005__pitch+0__seed0__PML.wav", "copied_to_this_report": false, "header": { "container": "RIFF/WAVE", "format_code": 1, "format": "PCM", "endianness": "little", "channels": 1, "sample_rate": 16000, "byte_rate": 32000, "block_align": 2, "bits_per_sample": 16, "sample_width_bytes": 2, "samples": 80000, "header_bytes": 44, "riff_size": 160036, "data_bytes": 160000, "header_hex": "524946462471020057415645666d74201000000001000100803e0000007d0000020010006461746100710200", "header_sha256": "fe37d7895050fe71672e015eccb2ba6a4ca09fed7bdcd9a838adf2c905a2ca42", "PCM_data_sha256": "cbc810aff3dbe4406947f341635acf81596653d56306d51ab8e21b19ed41167e" } }, { "case": "librispeech_1221-135766-0005", "pitch_semitones": 0, "seed": 0, "variant": "RAW_FULL", "bytes": 160044, "WAV_sha256": "c8d68c053cb8d282027b7de52953c8e2abf966d514b084763755213b8ca8482a", "source_NPY_sha256": "b45a6380b5c8f17818af0dfae734ac4982a0ee6a4490f8dfa38759fe2f4e2cb5", "samples": 80000, "sample_rate": 16000, "channels": 1, "source_WAV_sha256": "dfac116fef67209496318648ef9ae9377f86378ef65fc8b11f78f86a2453da34", "source_samples": 96000, "excerpt_sample_range": [ 0, 80000 ], "excerpt_seconds": [ 0, 5.0 ], "PCM_samples_unchanged": true, "cut_reason": "Publication route maximum five seconds", "sample_width_bytes": 2, "gain": 1, "normalization": "none", "clipping": "none", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/1221-135766-0005__pitch+0__seed0__RAW_FULL.wav", "copied_to_this_report": false, "header": { "container": "RIFF/WAVE", "format_code": 1, "format": "PCM", "endianness": "little", "channels": 1, "sample_rate": 16000, "byte_rate": 32000, "block_align": 2, "bits_per_sample": 16, "sample_width_bytes": 2, "samples": 80000, "header_bytes": 44, "riff_size": 160036, "data_bytes": 160000, "header_hex": "524946462471020057415645666d74201000000001000100803e0000007d0000020010006461746100710200", "header_sha256": "fe37d7895050fe71672e015eccb2ba6a4ca09fed7bdcd9a838adf2c905a2ca42", "PCM_data_sha256": "54b0af7761dde1a9d6d5b4af57803f12a92a6fb5959fb2e5d9de170558336b1b" } }, { "case": "librispeech_1089-134686-0032", "pitch_semitones": null, "seed": null, "variant": "ORIGINAL", "bytes": 130924, "WAV_sha256": "209ff141d5f38222d39b0c177d11271cb0fdf38ae0f03bb72305eb3f77543cab", "source_NPY_sha256": "cf1d3a15f9864fa36824ce1cb46c161624f17c4e82ba1e8b14109293167293b6", "samples": 65440, "sample_rate": 16000, "channels": 1, "source_WAV_sha256": "209ff141d5f38222d39b0c177d11271cb0fdf38ae0f03bb72305eb3f77543cab", "source_samples": 65440, "excerpt_sample_range": [ 0, 65440 ], "excerpt_seconds": [ 0, 4.09 ], "PCM_samples_unchanged": true, "cut_reason": "Full waveform within five seconds", "sample_width_bytes": 2, "gain": 1, "normalization": "none", "clipping": "none", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/1089-134686-0032__original.wav", "copied_to_this_report": false, "header": { "container": "RIFF/WAVE", "format_code": 1, "format": "PCM", "endianness": "little", "channels": 1, "sample_rate": 16000, "byte_rate": 32000, "block_align": 2, "bits_per_sample": 16, "sample_width_bytes": 2, "samples": 65440, "header_bytes": 44, "riff_size": 130916, "data_bytes": 130880, "header_hex": "5249464664ff010057415645666d74201000000001000100803e0000007d0000020010006461746140ff0100", "header_sha256": "40f6385e2c12909fb203eff75e76d10044107b36449e5cbae62b3f488a6b7bbe", "PCM_data_sha256": "643f787b905d5160255daf650b50e8a272ea88d166a6e7aca2a7ee357cf232fb" }, "reference_role": "LibriSpeech原音は内容・時刻の参考。音高編集音の正解音声ではない。" }, { "case": "librispeech_1089-134686-0032", "pitch_semitones": 0, "seed": 0, "variant": "PML", "bytes": 130924, "WAV_sha256": "c06b08e00da7f17b98a09810fec60b56da1e9d988a3e9ef7f7e8515965a82e1c", "source_NPY_sha256": "80bf274d7883bff30b22f1eae1ea2a40cb7d3f7fdc0bc75162dc90e44faa6cc8", "samples": 65440, "sample_rate": 16000, "channels": 1, "source_WAV_sha256": "c06b08e00da7f17b98a09810fec60b56da1e9d988a3e9ef7f7e8515965a82e1c", "source_samples": 65440, "excerpt_sample_range": [ 0, 65440 ], "excerpt_seconds": [ 0, 4.09 ], "PCM_samples_unchanged": true, "cut_reason": "Full waveform within five seconds", "sample_width_bytes": 2, "gain": 1, "normalization": "none", "clipping": "none", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/1089-134686-0032__pitch+0__seed0__PML.wav", "copied_to_this_report": false, "header": { "container": "RIFF/WAVE", "format_code": 1, "format": "PCM", "endianness": "little", "channels": 1, "sample_rate": 16000, "byte_rate": 32000, "block_align": 2, "bits_per_sample": 16, "sample_width_bytes": 2, "samples": 65440, "header_bytes": 44, "riff_size": 130916, "data_bytes": 130880, "header_hex": "5249464664ff010057415645666d74201000000001000100803e0000007d0000020010006461746140ff0100", "header_sha256": "40f6385e2c12909fb203eff75e76d10044107b36449e5cbae62b3f488a6b7bbe", "PCM_data_sha256": "233574f20b73ef1bebfd8b6779665b763e9ed069b2fde3de205e227165643bf7" } }, { "case": "librispeech_1089-134686-0032", "pitch_semitones": 0, "seed": 0, "variant": "RAW_FULL", "bytes": 130924, "WAV_sha256": "857563fdcff718426f63a7998c22ac42001835ed490b7ce35d245be6b5135fdf", "source_NPY_sha256": "949f626d7910d668b396b77782f8be4f4609bfac31d94289026258b8bb674406", "samples": 65440, "sample_rate": 16000, "channels": 1, "source_WAV_sha256": "857563fdcff718426f63a7998c22ac42001835ed490b7ce35d245be6b5135fdf", "source_samples": 65440, "excerpt_sample_range": [ 0, 65440 ], "excerpt_seconds": [ 0, 4.09 ], "PCM_samples_unchanged": true, "cut_reason": "Full waveform within five seconds", "sample_width_bytes": 2, "gain": 1, "normalization": "none", "clipping": "none", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/1089-134686-0032__pitch+0__seed0__RAW_FULL.wav", "copied_to_this_report": false, "header": { "container": "RIFF/WAVE", "format_code": 1, "format": "PCM", "endianness": "little", "channels": 1, "sample_rate": 16000, "byte_rate": 32000, "block_align": 2, "bits_per_sample": 16, "sample_width_bytes": 2, "samples": 65440, "header_bytes": 44, "riff_size": 130916, "data_bytes": 130880, "header_hex": "5249464664ff010057415645666d74201000000001000100803e0000007d0000020010006461746140ff0100", "header_sha256": "40f6385e2c12909fb203eff75e76d10044107b36449e5cbae62b3f488a6b7bbe", "PCM_data_sha256": "61f65f443ada9799b6670cfe4af55ebdbb1381a64deb5be6502188547c56e584" } }, { "case": "librispeech_121-121726-0012", "pitch_semitones": -3, "seed": 0, "variant": "PML", "bytes": 129484, "WAV_sha256": "18e7dd0140ad704e8dc15069a7e7466d77983d1a8e0bd32aecdde1a983e5b53f", "source_NPY_sha256": "b5d58d1a3452c7cced18f708f3b23c109acb016e90f0ab159606d447fa2a8ba2", "samples": 64720, "sample_rate": 16000, "channels": 1, "source_WAV_sha256": "18e7dd0140ad704e8dc15069a7e7466d77983d1a8e0bd32aecdde1a983e5b53f", "source_samples": 64720, "excerpt_sample_range": [ 0, 64720 ], "excerpt_seconds": [ 0, 4.045 ], "PCM_samples_unchanged": true, "cut_reason": "Full waveform within five seconds", "sample_width_bytes": 2, "gain": 1, "normalization": "none", "clipping": "none", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/121-121726-0012__pitch-3__seed0__PML.wav", "copied_to_this_report": false, "header": { "container": "RIFF/WAVE", "format_code": 1, "format": "PCM", "endianness": "little", "channels": 1, "sample_rate": 16000, "byte_rate": 32000, "block_align": 2, "bits_per_sample": 16, "sample_width_bytes": 2, "samples": 64720, "header_bytes": 44, "riff_size": 129476, "data_bytes": 129440, "header_hex": "52494646c4f9010057415645666d74201000000001000100803e0000007d00000200100064617461a0f90100", "header_sha256": "1430049218770f1f842b7d92f13fc6a86ed5691e022e91f9ce5ddc8397e15f73", "PCM_data_sha256": "5fb1d745ea987235c3a02750bbd36f3c28e418fc9f45d5e3714f1104f9c329dd" } }, { "case": "librispeech_121-121726-0012", "pitch_semitones": -3, "seed": 0, "variant": "RAW_SHAPE", "bytes": 129484, "WAV_sha256": "3175a70708b6ad13c0af64cf626fda45f0837fa1d2807095b3bdb60c643823e3", "source_NPY_sha256": "758b0ed995d72ee4715440ce181c32f90ce018986d8010890368718a09bae037", "samples": 64720, "sample_rate": 16000, "channels": 1, "source_WAV_sha256": "3175a70708b6ad13c0af64cf626fda45f0837fa1d2807095b3bdb60c643823e3", "source_samples": 64720, "excerpt_sample_range": [ 0, 64720 ], "excerpt_seconds": [ 0, 4.045 ], "PCM_samples_unchanged": true, "cut_reason": "Full waveform within five seconds", "sample_width_bytes": 2, "gain": 1, "normalization": "none", "clipping": "none", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/121-121726-0012__pitch-3__seed0__RAW_SHAPE.wav", "copied_to_this_report": false, "header": { "container": "RIFF/WAVE", "format_code": 1, "format": "PCM", "endianness": "little", "channels": 1, "sample_rate": 16000, "byte_rate": 32000, "block_align": 2, "bits_per_sample": 16, "sample_width_bytes": 2, "samples": 64720, "header_bytes": 44, "riff_size": 129476, "data_bytes": 129440, "header_hex": "52494646c4f9010057415645666d74201000000001000100803e0000007d00000200100064617461a0f90100", "header_sha256": "1430049218770f1f842b7d92f13fc6a86ed5691e022e91f9ce5ddc8397e15f73", "PCM_data_sha256": "7f02eb79c1df3adf6917721fba9e350769d61b7cf6bd8bc25e59d28c5d1264f2" } }, { "case": "librispeech_121-121726-0012", "pitch_semitones": -3, "seed": 0, "variant": "RAW_FULL", "bytes": 129484, "WAV_sha256": "8d8f880ea646a1a25e7555afdf936d47ac4371be2e538c12cb41de2a33540666", "source_NPY_sha256": "c8da7838eee5e846ea1ea3e638a4f7270f7be433fc9c893642421fd34bbe512e", "samples": 64720, "sample_rate": 16000, "channels": 1, "source_WAV_sha256": "8d8f880ea646a1a25e7555afdf936d47ac4371be2e538c12cb41de2a33540666", "source_samples": 64720, "excerpt_sample_range": [ 0, 64720 ], "excerpt_seconds": [ 0, 4.045 ], "PCM_samples_unchanged": true, "cut_reason": "Full waveform within five seconds", "sample_width_bytes": 2, "gain": 1, "normalization": "none", "clipping": "none", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/121-121726-0012__pitch-3__seed0__RAW_FULL.wav", "copied_to_this_report": false, "header": { "container": "RIFF/WAVE", "format_code": 1, "format": "PCM", "endianness": "little", "channels": 1, "sample_rate": 16000, "byte_rate": 32000, "block_align": 2, "bits_per_sample": 16, "sample_width_bytes": 2, "samples": 64720, "header_bytes": 44, "riff_size": 129476, "data_bytes": 129440, "header_hex": "52494646c4f9010057415645666d74201000000001000100803e0000007d00000200100064617461a0f90100", "header_sha256": "1430049218770f1f842b7d92f13fc6a86ed5691e022e91f9ce5ddc8397e15f73", "PCM_data_sha256": "9a72aae85cf058661692ce23a54e67173cc445770c320a18d05de53ec62d0bb4" } }, { "case": "librispeech_121-121726-0012", "pitch_semitones": -3, "seed": 20261015, "variant": "PML", "bytes": 129484, "WAV_sha256": "169de85768a72963c76c17815f3ab61a18e40b1c3057e5a63ef39e972795fd71", "source_NPY_sha256": "2ed85deb832fe249f779569e7c6379b57d3db0525df04a26444f4d5151db0b89", "samples": 64720, "sample_rate": 16000, "channels": 1, "source_WAV_sha256": "169de85768a72963c76c17815f3ab61a18e40b1c3057e5a63ef39e972795fd71", "source_samples": 64720, "excerpt_sample_range": [ 0, 64720 ], "excerpt_seconds": [ 0, 4.045 ], "PCM_samples_unchanged": true, "cut_reason": "Full waveform within five seconds", "sample_width_bytes": 2, "gain": 1, "normalization": "none", "clipping": "none", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/121-121726-0012__pitch-3__seed20261015__PML.wav", "copied_to_this_report": false, "header": { "container": "RIFF/WAVE", "format_code": 1, "format": "PCM", "endianness": "little", "channels": 1, "sample_rate": 16000, "byte_rate": 32000, "block_align": 2, "bits_per_sample": 16, "sample_width_bytes": 2, "samples": 64720, "header_bytes": 44, "riff_size": 129476, "data_bytes": 129440, "header_hex": "52494646c4f9010057415645666d74201000000001000100803e0000007d00000200100064617461a0f90100", "header_sha256": "1430049218770f1f842b7d92f13fc6a86ed5691e022e91f9ce5ddc8397e15f73", "PCM_data_sha256": "a95177fa38b9e93b01531643790261dd8b44e9fd40118a2a6c6dfa885f52860e" } }, { "case": "librispeech_121-121726-0012", "pitch_semitones": -3, "seed": 20261015, "variant": "RAW_SHAPE", "bytes": 129484, "WAV_sha256": "2a8f46ed0459425f8fecadae69f02ead56f4b91935bab6013c8ea33c2b320cc9", "source_NPY_sha256": "efd7cd6a2dca10baebb8f1cc44f3bf043cb8e0ead5c5927beb0d0ccf208f03f7", "samples": 64720, "sample_rate": 16000, "channels": 1, "source_WAV_sha256": "2a8f46ed0459425f8fecadae69f02ead56f4b91935bab6013c8ea33c2b320cc9", "source_samples": 64720, "excerpt_sample_range": [ 0, 64720 ], "excerpt_seconds": [ 0, 4.045 ], "PCM_samples_unchanged": true, "cut_reason": "Full waveform within five seconds", "sample_width_bytes": 2, "gain": 1, "normalization": "none", "clipping": "none", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/121-121726-0012__pitch-3__seed20261015__RAW_SHAPE.wav", "copied_to_this_report": false, "header": { "container": "RIFF/WAVE", "format_code": 1, "format": "PCM", "endianness": "little", "channels": 1, "sample_rate": 16000, "byte_rate": 32000, "block_align": 2, "bits_per_sample": 16, "sample_width_bytes": 2, "samples": 64720, "header_bytes": 44, "riff_size": 129476, "data_bytes": 129440, "header_hex": "52494646c4f9010057415645666d74201000000001000100803e0000007d00000200100064617461a0f90100", "header_sha256": "1430049218770f1f842b7d92f13fc6a86ed5691e022e91f9ce5ddc8397e15f73", "PCM_data_sha256": "1caad9167627320eeff2dd5a95ef4183b2e09c7331768617116d7bcd80b36f85" } }, { "case": "librispeech_121-121726-0012", "pitch_semitones": -3, "seed": 20261015, "variant": "RAW_FULL", "bytes": 129484, "WAV_sha256": "617484725c170f40b86cefca5a42ab00ae91a455916e31cc4bd4b31f06ac241c", "source_NPY_sha256": "5301869c730597ada086b2a0b7274654e22a93872364e3a7868e1d74098902c2", "samples": 64720, "sample_rate": 16000, "channels": 1, "source_WAV_sha256": "617484725c170f40b86cefca5a42ab00ae91a455916e31cc4bd4b31f06ac241c", "source_samples": 64720, "excerpt_sample_range": [ 0, 64720 ], "excerpt_seconds": [ 0, 4.045 ], "PCM_samples_unchanged": true, "cut_reason": "Full waveform within five seconds", "sample_width_bytes": 2, "gain": 1, "normalization": "none", "clipping": "none", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/121-121726-0012__pitch-3__seed20261015__RAW_FULL.wav", "copied_to_this_report": false, "header": { "container": "RIFF/WAVE", "format_code": 1, "format": "PCM", "endianness": "little", "channels": 1, "sample_rate": 16000, "byte_rate": 32000, "block_align": 2, "bits_per_sample": 16, "sample_width_bytes": 2, "samples": 64720, "header_bytes": 44, "riff_size": 129476, "data_bytes": 129440, "header_hex": "52494646c4f9010057415645666d74201000000001000100803e0000007d00000200100064617461a0f90100", "header_sha256": "1430049218770f1f842b7d92f13fc6a86ed5691e022e91f9ce5ddc8397e15f73", "PCM_data_sha256": "ee7e0ab25478fcc8d31158232de0e77a43e7dd21d4c8e1adc286303e64ae4a12" } }, { "case": "librispeech_121-121726-0012", "pitch_semitones": 3, "seed": 0, "variant": "PML", "bytes": 129484, "WAV_sha256": "d1fcaee4b6a8981f09e6b6b8583ad741e7ceaeb766e1a8b913b64f13dc88746a", "source_NPY_sha256": "082bd02b2d99751a844f503d91a3c6df5c7f98d60aef01c2c8d9ee481295c776", "samples": 64720, "sample_rate": 16000, "channels": 1, "source_WAV_sha256": "d1fcaee4b6a8981f09e6b6b8583ad741e7ceaeb766e1a8b913b64f13dc88746a", "source_samples": 64720, "excerpt_sample_range": [ 0, 64720 ], "excerpt_seconds": [ 0, 4.045 ], "PCM_samples_unchanged": true, "cut_reason": "Full waveform within five seconds", "sample_width_bytes": 2, "gain": 1, "normalization": "none", "clipping": "none", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/121-121726-0012__pitch+3__seed0__PML.wav", "copied_to_this_report": false, "header": { "container": "RIFF/WAVE", "format_code": 1, "format": "PCM", "endianness": "little", "channels": 1, "sample_rate": 16000, "byte_rate": 32000, "block_align": 2, "bits_per_sample": 16, "sample_width_bytes": 2, "samples": 64720, "header_bytes": 44, "riff_size": 129476, "data_bytes": 129440, "header_hex": "52494646c4f9010057415645666d74201000000001000100803e0000007d00000200100064617461a0f90100", "header_sha256": "1430049218770f1f842b7d92f13fc6a86ed5691e022e91f9ce5ddc8397e15f73", "PCM_data_sha256": "29e50e42612e9e526e6f71a812f316d446b867818a35bb28d88f204eed4f6992" } }, { "case": "librispeech_121-121726-0012", "pitch_semitones": 3, "seed": 0, "variant": "RAW_SHAPE", "bytes": 129484, "WAV_sha256": "7145fde331940ff1549669e59f5f1aa5003234a0d08234d5c9bd5b92818727aa", "source_NPY_sha256": "1ee658832e1ddbc0fead3358ce91f9abb4c6243747772c1b1995ed0390499693", "samples": 64720, "sample_rate": 16000, "channels": 1, "source_WAV_sha256": "7145fde331940ff1549669e59f5f1aa5003234a0d08234d5c9bd5b92818727aa", "source_samples": 64720, "excerpt_sample_range": [ 0, 64720 ], "excerpt_seconds": [ 0, 4.045 ], "PCM_samples_unchanged": true, "cut_reason": "Full waveform within five seconds", "sample_width_bytes": 2, "gain": 1, "normalization": "none", "clipping": "none", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/121-121726-0012__pitch+3__seed0__RAW_SHAPE.wav", "copied_to_this_report": false, "header": { "container": "RIFF/WAVE", "format_code": 1, "format": "PCM", "endianness": "little", "channels": 1, "sample_rate": 16000, "byte_rate": 32000, "block_align": 2, "bits_per_sample": 16, "sample_width_bytes": 2, "samples": 64720, "header_bytes": 44, "riff_size": 129476, "data_bytes": 129440, "header_hex": "52494646c4f9010057415645666d74201000000001000100803e0000007d00000200100064617461a0f90100", "header_sha256": "1430049218770f1f842b7d92f13fc6a86ed5691e022e91f9ce5ddc8397e15f73", "PCM_data_sha256": "744ffcac0d6246b7ec3f8464e1f044ab9d62209e6d71a0b8c424e6c20175cd52" } }, { "case": "librispeech_121-121726-0012", "pitch_semitones": 3, "seed": 0, "variant": "RAW_FULL", "bytes": 129484, "WAV_sha256": "8b97a060d79d1f74bee0a6cd3f3e267d400be0c988b5890eba436bddcc21ccee", "source_NPY_sha256": "448a41231e539e37d40953699caf50f78ed0866197fa9c0aa01c689ca773f9f8", "samples": 64720, "sample_rate": 16000, "channels": 1, "source_WAV_sha256": "8b97a060d79d1f74bee0a6cd3f3e267d400be0c988b5890eba436bddcc21ccee", "source_samples": 64720, "excerpt_sample_range": [ 0, 64720 ], "excerpt_seconds": [ 0, 4.045 ], "PCM_samples_unchanged": true, "cut_reason": "Full waveform within five seconds", "sample_width_bytes": 2, "gain": 1, "normalization": "none", "clipping": "none", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/121-121726-0012__pitch+3__seed0__RAW_FULL.wav", "copied_to_this_report": false, "header": { "container": "RIFF/WAVE", "format_code": 1, "format": "PCM", "endianness": "little", "channels": 1, "sample_rate": 16000, "byte_rate": 32000, "block_align": 2, "bits_per_sample": 16, "sample_width_bytes": 2, "samples": 64720, "header_bytes": 44, "riff_size": 129476, "data_bytes": 129440, "header_hex": "52494646c4f9010057415645666d74201000000001000100803e0000007d00000200100064617461a0f90100", "header_sha256": "1430049218770f1f842b7d92f13fc6a86ed5691e022e91f9ce5ddc8397e15f73", "PCM_data_sha256": "7ddf3c724874bee5c46f1ce7dab942633d4b8ec483df215533ec7694d4953433" } }, { "case": "librispeech_121-121726-0012", "pitch_semitones": 3, "seed": 20261015, "variant": "PML", "bytes": 129484, "WAV_sha256": "1e815a8c64819cc373635ee742c8c0debaa1934fddd47b183d3769ccff0e61f3", "source_NPY_sha256": "d56d4907bb8ee291d7ca7ff2c240ef4a1bdb18d485c78449b45dd152818c0e26", "samples": 64720, "sample_rate": 16000, "channels": 1, "source_WAV_sha256": "1e815a8c64819cc373635ee742c8c0debaa1934fddd47b183d3769ccff0e61f3", "source_samples": 64720, "excerpt_sample_range": [ 0, 64720 ], "excerpt_seconds": [ 0, 4.045 ], "PCM_samples_unchanged": true, "cut_reason": "Full waveform within five seconds", "sample_width_bytes": 2, "gain": 1, "normalization": "none", "clipping": "none", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/121-121726-0012__pitch+3__seed20261015__PML.wav", "copied_to_this_report": false, "header": { "container": "RIFF/WAVE", "format_code": 1, "format": "PCM", "endianness": "little", "channels": 1, "sample_rate": 16000, "byte_rate": 32000, "block_align": 2, "bits_per_sample": 16, "sample_width_bytes": 2, "samples": 64720, "header_bytes": 44, "riff_size": 129476, "data_bytes": 129440, "header_hex": "52494646c4f9010057415645666d74201000000001000100803e0000007d00000200100064617461a0f90100", "header_sha256": "1430049218770f1f842b7d92f13fc6a86ed5691e022e91f9ce5ddc8397e15f73", "PCM_data_sha256": "7afa7f2c9146ad47969ea4c5969f7db28e7cb1515689e02a7f4f36dec46bc560" } }, { "case": "librispeech_121-121726-0012", "pitch_semitones": 3, "seed": 20261015, "variant": "RAW_SHAPE", "bytes": 129484, "WAV_sha256": "345c2114858447d84db19479327b3d9e96bfd25c77f29d668c161a791c0f8433", "source_NPY_sha256": "4f8dc9a31dba23758b8ba40719eee80f25a3ac6ad731fa319bf29302288eac6c", "samples": 64720, "sample_rate": 16000, "channels": 1, "source_WAV_sha256": "345c2114858447d84db19479327b3d9e96bfd25c77f29d668c161a791c0f8433", "source_samples": 64720, "excerpt_sample_range": [ 0, 64720 ], "excerpt_seconds": [ 0, 4.045 ], "PCM_samples_unchanged": true, "cut_reason": "Full waveform within five seconds", "sample_width_bytes": 2, "gain": 1, "normalization": "none", "clipping": "none", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/121-121726-0012__pitch+3__seed20261015__RAW_SHAPE.wav", "copied_to_this_report": false, "header": { "container": "RIFF/WAVE", "format_code": 1, "format": "PCM", "endianness": "little", "channels": 1, "sample_rate": 16000, "byte_rate": 32000, "block_align": 2, "bits_per_sample": 16, "sample_width_bytes": 2, "samples": 64720, "header_bytes": 44, "riff_size": 129476, "data_bytes": 129440, "header_hex": "52494646c4f9010057415645666d74201000000001000100803e0000007d00000200100064617461a0f90100", "header_sha256": "1430049218770f1f842b7d92f13fc6a86ed5691e022e91f9ce5ddc8397e15f73", "PCM_data_sha256": "a9a5bb2c179054d1430ae6497d2891043ed82180a9a573d89389870bac596e54" } }, { "case": "librispeech_121-121726-0012", "pitch_semitones": 3, "seed": 20261015, "variant": "RAW_FULL", "bytes": 129484, "WAV_sha256": "95a05df2003897deac9c19b0cbc778bf928136f828748add8113707d6a8ec2c9", "source_NPY_sha256": "53a9de0adae6d1f1b59307c0e9d5345f9cbdd7fc155be3ba20e9327d5ae7085b", "samples": 64720, "sample_rate": 16000, "channels": 1, "source_WAV_sha256": "95a05df2003897deac9c19b0cbc778bf928136f828748add8113707d6a8ec2c9", "source_samples": 64720, "excerpt_sample_range": [ 0, 64720 ], "excerpt_seconds": [ 0, 4.045 ], "PCM_samples_unchanged": true, "cut_reason": "Full waveform within five seconds", "sample_width_bytes": 2, "gain": 1, "normalization": "none", "clipping": "none", "url": "https://hiroshiba.github.io/check-nonml-saikyo-vocoder/research/raw-envelope-risk-2026-10-19/audio/121-121726-0012__pitch+3__seed20261015__RAW_FULL.wav", "copied_to_this_report": false, "header": { "container": "RIFF/WAVE", "format_code": 1, "format": "PCM", "endianness": "little", "channels": 1, "sample_rate": 16000, "byte_rate": 32000, "block_align": 2, "bits_per_sample": 16, "sample_width_bytes": 2, "samples": 64720, "header_bytes": 44, "riff_size": 129476, "data_bytes": 129440, "header_hex": "52494646c4f9010057415645666d74201000000001000100803e0000007d00000200100064617461a0f90100", "header_sha256": "1430049218770f1f842b7d92f13fc6a86ed5691e022e91f9ce5ddc8397e15f73", "PCM_data_sha256": "d595fc6a0771bdcd0aaa6098e2b1d33ded4f260f832e8dbc0ac37f167dabebcb" } } ] }