This is so cool. Yes, Melodyne has been around for a while, saving vocalists from their own flatness. But what really makes this interesting is what it means for samples. It’s hard enough now to detect the use of a sample, as it is folded, spun, and mutilated. But at what point does it stop being a sample. Obviously, the phrasing of the original instrument remains important, but at some point, you’ve chopped the original up so much that it isn’t really a “sample” any more at all?
I mean, if you “sample” a paragraph from a book, and don’t cite it, you’ve clearly plagiarized, but when you take single words from someone, you clearly haven’t. Musical notes are not as discrete as words–the phrasing of a note by a violinist, or even by a pianist is fairly unique–but once you get down to the note-by-note level, it starts to feel a lot less like sampling.