[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-79193":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":9,"language":10,"languages":9,"totalLinesOfCode":9,"stars":11,"forks":12,"watchers":13,"openIssues":14,"contributorsCount":14,"subscribersCount":14,"size":14,"stars1d":14,"stars7d":14,"stars30d":15,"stars90d":14,"forks30d":14,"starsTrendScore":14,"compositeScore":16,"rankGlobal":9,"rankLanguage":9,"license":17,"archived":18,"fork":18,"defaultBranch":19,"hasWiki":20,"hasPages":18,"topics":21,"createdAt":9,"pushedAt":9,"updatedAt":22,"readmeContent":23,"aiSummary":24,"trendingCount":14,"starSnapshotCount":14,"syncStatus":25,"lastSyncTime":26,"discoverSource":27},79193,"tts-prosody-probe","saadansha\u002Ftts-prosody-probe","saadansha","Probe and compare the prosody (pitch \u002F energy \u002F duration) of TTS outputs.",null,"Python",223,11459,6,0,189,60,"Other",false,"main",true,[],"2026-06-12 04:01:24","# tts-prosody-probe\n\nA tiny CLI \u002F library to extract and compare the **prosody** (pitch, energy,\nduration) of TTS audio outputs. Use it when you want to know whether your\nnew TTS model actually changed how natural a sentence *sounds*, beyond\nspectral fidelity.\n\n## Install\n\n```\npip install tts-prosody-probe\n```\n\n(System dep: `libsndfile`. On Ubuntu: `apt install libsndfile1`.)\n\n## Usage\n\n```python\nfrom tpp import pitch_contour, compare_pair\n\nf0 = pitch_contour(\"a.wav\")             # numpy array (Hz, NaN for unvoiced)\ndiff = compare_pair(\"a.wav\", \"b.wav\")   # {\"f0_rmse\", \"energy_rmse\", \"dur_ratio\"}\n```\n\nCLI:\n\n```bash\ntpp extract a.wav --out a.npz\ntpp compare a.wav b.wav\ntpp plot a.wav --out a.png\n```\n\n## What you get\n\n| Metric           | What it measures                                        |\n|------------------|---------------------------------------------------------|\n| `f0_rmse`        | RMSE of pitch contour (Hz) over voiced frames           |\n| `f0_corr`        | Pearson correlation of voiced-frame pitch contours      |\n| `energy_rmse`    | RMSE of frame-level RMS energy                          |\n| `dur_ratio`      | duration(B) \u002F duration(A) — how much longer the output is |\n| `voicing_iou`    | IoU of voiced-frame masks                               |\n\nUse `tpp compare --metric f0_corr` to focus on a single number.\n\n## License\n\nGPL-3.0.\n","该项目是一个用于提取和比较TTS音频输出的韵律（音高、能量、时长）的小型CLI工具或库。其核心功能包括通过Python接口或命令行界面提取音频文件的音高轮廓，并对比两个音频文件之间的韵律差异，提供如音高RMSE、能量RMSE等指标来量化这些差异。适合于评估新的TTS模型是否在自然度上有所改进，而不仅仅是频谱保真度上的变化。对于从事语音合成研究与开发的专业人士来说，这是一个实用的辅助工具。",2,"2026-06-06 03:58:39","CREATED_QUERY"]