Can this repo clone the original voice and generate a voice file with the speaker voice ?

[misraj-ah]

I'm a little bit confused about what can this repo really do, the readme file says that we can clone the voice, but in issues I found that this repo can only clone the tone color of the speaker, and I don't know what is this exactly mean ?

[johnwick123f]

Yes it can clone voices. What this repo means by tone color is that emotion or volume isn't really converted but rather the actual style. This codebase works by a tts model generating speech and a voice converter to make it sound like your speaker you want to clone. Emotion and volume is controlled by the tts model and you can actually swap that out.

[misraj-ah]

Thank for your reply @johnwick123f

Could you please tell me what is the name of the model that clone the voice ? The ToneColorConvertor model generates voices with a completely different voice