I was playing with some loops of Tuvan Throat Singing a couple of nights ago, and thought what a shame it is that this artform is only ever represented by a few loops, and nothing more flexible.
With Gary's Sonic-Morphing now a reality I thought we could see a virtual Khoomei artist. Say you made 'static' samples of a sustained fundamental, with every variation of partials that the singer is capable of producing. Also transition samples between 'adjacent' sets of partials. When the mod wheel is pushed the current set of partials could be sonically-morphed into a transition sample, which could then be morphed into the 'static' sample of the fundamental with the next highest partial being prominent.
Of course the whole of this process would have to be do-able over several vowel sounds, which could be switched using key-switches. And there'd have to be consonants, with some kind of word-building utility. And there would have to be release samples of the fundamental 'collapsing' during every set of partials. Also the whole thing should be done with different fundamentals.
Re: Virtual Throat Singer - a challenge to developers
Actually, you might want to talk to Audionerds, they've already got a lot of the processes (though that's just a filter.) Western vocals are comlicated enough with only one note at a time, to learn to play Khoomei from a keyboard might take a bit of work,,,
------- It's all about the music - really. I keep telling myself that...