Originally Posted By: SongCrafters
Do you think they will ever achieve true human-sounding vocals?

You'll know it when you hear it. wink

How about this? The vocal is a bit back in the mix, but unless someone told me, I wouldn't have thought the vocal was synthetic:



Then again, since I don't speak Japanese, I'm not as attuned to the discrepancies which - when I hear them in the English version - seem to stick out like a sore thumb.

Here's the catch: doing this well is pretty much the same as putting together a convincing MIDI guitar solo. You can't just drop notes into the editor, you have to know how to shape notes and phrases like a singer. You need to make decisions a singer would made, often on a microscopic scale.

That's not an trivial skill, and it can take a lot of time to build a performance. That's time that could be saved by having a competent singer perform the track instead.

That said, they're also working on training the programs to learn and automatically apply vocal styles.

-- David Cuny
My virtual singer development blog

Vocal control, you say. Never heard of it. Is that some kind of ProTools thing?