Currently there exists text to speech, and vocal to vocal.

There is not currently any method by which to create a song with original lyrics.

It looks like VoiceMod is developing that ability already, I would love for them to take it all the way...

Make a standalone, unlimited use, offline, CUDA powered, simple to use, easy to install, text to song.

I imagine it as such:

A: Choose song to cover, input (or auto detect) original song lyrics

B: Choose voice to cover, input parody song lyrics

C: Maybe some kind of checking to make sure you have the new lyrics lined up over the original?

That's all I wanna be able to do. I should be able to take any song to use as the base, and input any lyrics with any RVC voice model to do the cover.

Please?