Clarify realtime voice reference use
This commit is contained in:
parent
2e941e409c
commit
e38f0c589e
3 changed files with 78 additions and 0 deletions
|
|
@ -72,3 +72,27 @@ It may include:
|
|||
- `delivery_pressure`
|
||||
|
||||
This is a request to AquaSynth, not a rendered audio contract.
|
||||
|
||||
## Reference Audio Artifacts
|
||||
|
||||
Realtime voice outputs may be attached as reference artifacts for tuning.
|
||||
|
||||
Suggested fields:
|
||||
|
||||
- `reference_audio.provider`
|
||||
- `reference_audio.voice`
|
||||
- `reference_audio.prompt_ref`
|
||||
- `reference_audio.transcript`
|
||||
- `reference_audio.audio_ref`
|
||||
- `reference_audio.rating_notes`
|
||||
- `reference_audio.allowed_use`
|
||||
|
||||
Allowed use should distinguish:
|
||||
|
||||
- `style_reference`
|
||||
- `prosody_reference`
|
||||
- `regression_reference`
|
||||
- `physical_ground_truth_forbidden`
|
||||
|
||||
Reference audio must not replace `phonetic_intent_request` or AquaSynth's
|
||||
articulatory reports.
|
||||
|
|
|
|||
Loading…
Reference in a new issue