# Voice Rights Summary

Speaker ID: `teacher-liu-authorized`

Status: authorized by user report.

Important files:

- Consent record: `storage/voice-rights/teacher-liu-authorized/consent.md`
- Source audio: `storage/voice-rights/teacher-liu-authorized/source-audio/reference-youtube-v2EN9GVEeMc.m4a`
- Cleaned WAV: `storage/voice-rights/teacher-liu-authorized/cleaned-audio/reference-youtube-v2EN9GVEeMc-clean.wav`
- Training manifest: `storage/voice-rights/teacher-liu-authorized/training-manifest.json`
- Voice model config: `storage/voice-rights/teacher-liu-authorized/voice-model.json`

Production rule:

- Use authorized voice audio if per-segment MP3s exist in:

```text
storage/voice-rights/teacher-liu-authorized/generated-audio/
```

- Use Edge-TTS fallback only if provider output is unavailable or not demo-ready.

Run with authorized audio:

```bash
AUTHORIZED_VOICE_SPEAKER_ID=teacher-liu-authorized \
npm run video:childhood-wonder -w @mentorai/api
```

Run with authorized audio required:

```bash
AUTHORIZED_VOICE_SPEAKER_ID=teacher-liu-authorized \
AUTHORIZED_VOICE_REQUIRED=true \
npm run video:childhood-wonder -w @mentorai/api
```

Before commercial publication:

- Replace operational consent note with signed consent, email confirmation, or archived screenshot.
- Confirm whether likeness/name usage is also authorized.