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?
