I know there's another AI related thread here, but it's mainly about text to speech.
This thread is to discuss the potential of voice cloning. Think about it. If you have gotten addicted to the voice of one hypnotist, you could clone their voice and use it on other files so that you effectively have endless content, espescially if they stop making content.
I recently tried this with Miss Lilith and had interesting resulst. I used RVC to train a model of her voice from a 9 minute recording. I only did 250 epochs, which took like 20-30 minutes.
I then used a random hypno recording (a Madeira premie clip) to see if it could get Lilith's voice.
I've included both the original file and the Lilith cloned audio.
Here's what the results sound like. aHR0cHM6Ly9maWxlcy5mbS91L241cmR2bmNxNXQ=
Of course, the biggest problem with cloning is that most hypno files have a lot of effects and echoes and im not sure how RVC will handle it.
It did do a decent job with some of Ҝ卂丂卄卂 丂卄卂Ҝㄒ丨's pieces, but I haven't tested it much. I feel like this would work really well for tists who don't use too many sfx and shit.
Have you guys tried this suff out? any tips or advice? Looking forward to hearing what you think of this.