What's the best way to clone your voice?

[Connor56]

I realise Thorsten's tutorial is recommended as the best place to start looking at cloning. I've watched his video, and it's good but I would like to dig deeper into how training VITS actually works.

Does anyone know of a more thorough tutorial for this? In particular:

• How long voice samples can be,
• How aligned they have to be with the text,
• The sort of punctuation that's supported,
• The computer specs you need to train efficiently,
• The amount of audio you need to get a good clone,
• What checkpoints to use for what voice,
• More direct information on the dataset and cloning procedure.

For anyone interested, this was a useful video I found on voice cloning for StyleTTS:

https://www.youtube.com/watch?v=5-Dk3ooxn2Q

If this doesn't exist, I will figure this out myself and post an article + video tutorial in the next couple of months.