It’s funny that while you’re still hoping that Dreamtonics will provide the ability to create your own AI voicebanks in some indefinite future and you find arguments that it’s impossible, in the meantime the community has created a program that does exactly what you want. It’s called DiffSinger and is available for free as open source on github. Of course, using it requires some technical knowledge and hardware resources, but if someone is really determined to have a private AI vocalist, nothing stands in the way in 2024.
「いいね！」 1
@green I already use RVC, it’s also for free… But, why should we not dream? We are with Dreamtonics, we are the Tone Dreamers… 
「いいね！」 1
I have a concept for a SynthV Studio Voicebank that probably won’t become official, but it’s pretty good tbh in my opinion: https://scratch.mit.edu/projects/946094419
「いいね！」 2
@BellaClowncore16 Interesting. As I don’t have any experience with scratch I don’t what I’m missing here, but I get an error in both Firefox and Chrome when checking out the project… The box on the left shows “Oops? Something went wrong. We are so sorry…”
I’m not sure why you’re directing people to a Scratch program, since all it does is show an image. Dreamtonics generally doesn’t have graphics for their characters.
You could include simply paste it into your post:
Voicebank Concept:
- Sounds like a soft-sounding young girl who’s british
- Japanese and English voicebanks
- AI Voicebank
「いいね！」 3