Emotion Transfer: perfomance mode

Emotion Transfer extracts the original performance and recreates it with any supported voice.

It helps reproduce the original delivery — including intonation, intensity, rhythm, and timing — while keeping the selected target voice.

Compatibility

Languages	English (US), Spanish (Castilian), Spanish (Latin American), Brazilian Portuguese, French (Parisian), German, Russian
S-tier, A-tier voices	Yes
Expressive, Standard clone	Yes
Voice design	Yes
B-tier, C-tier voices	No
Voice acting	No
Prompts	Not yet supported, works as an alternative

Limitations

Emotion Transfer is not compatible with prompts or Voice Acting.

It works as an alternative way to control performance. When Emotion Transfer is used, the chunk performance is based on the selected reference chunk instead of manual prompts or recorded voice acting.

Apply Emotion Transfer to a voice

Open a Lip-sync task.
Select a chunk with an assigned voice.

Move the performance control slider to Emotion Transfer.

Wait for the first two generated versions.
- Emotion — the extracted performance from the original chunk. Listen it to review the transferred intonation, intensity, etc.
- Voice — the selected target voice with the transferred performance applied.
Click Generate more versions if necessary.
Click Use to apply the preferred version to the chunk.
Save progress.

Save versions for comparison

Click Use on any version that should be kept for later comparison.

Each selected version is added to Saved versions. This makes it possible to keep several good options and choose the best one before processing the final audio.

Fine-tune Emotion Transfer

Use the advanced controls to adjust how the original performance is transferred.

Change chunk reference	Select a different original chunk as the performance reference.
Intonation	Controls how much the pitch rises and falls. Lower values sound more even; higher values sound more expressive.
Intensity	Controls how much vocal energy is used. Lower values sound softer and gentler; higher values sound louder and more intense.
Variation	Controls how much the voice can vary from its natural tone. Lower values keep it closer to its original sound; higher values allow more varied performance.
Duration	Adjusts the target audio duration without changing playback speed. This helps make speech naturally faster or slower while preserving audio quality.