If I can make my voice sound like Kanye from a 10-minute sample of Kanye, then why couldn't I get my own voice recorded badly to sound like my own voice recorded well from a good 10-minute sample? What is the best tool/website/method for this kind of thing?
Suppose someone was wearing a lav microphone that was recorded perfectly the first half of the day but was slightly muffled/quieter the second half of the day on a documentary video shoot. Could a good sample of clean audio that we have from the first half of the day be uploaded to an AI tool to have this person's voice figured out, and then use that to recreate and replace some audio from the second half of the day where it was muffled/quieter?
[link] [comments]