Voice Changer takes any audio of someone speaking and re-records it in a different voice — keeping the original pauses, emotion, pitch contour and emphasis.
When to use it
- You recorded a great take but want a different voice.
- You want to anonymise a real interview.
- You're prototyping a character and want to test multiple voices on the same performance.
Source quality matters
The model preserves performance. If the source is mumbly, the output is mumbly. If the source is well-recorded with clear diction, the output sings.
Preserve original style
At 0, you get a clean delivery in the target voice. At 1, you keep heavy inflection from the source. Default 0.6 is the sweet spot — character without colour-bleed.
Remove background noise
Always on, unless your source is already studio-clean.