Talk:Hatsune Miku/@comment-4670969-20150212163502/@comment-82.131.232.19-20150221174500

I think Vocaloid5 technology will be about CFD: Computational Fluid Dynamics. This would mean Vocaloid5 partly returns to the Vocaloid1 era, in that voice synthesis will not simply happen via combining the atomic sound samples provided by a human donor, e.g. Saki Fujita san.

Just like V1, V5 may also include numerical simulation of the interaction between air and human physiology, for example making Miku able to sing better then Saki Fujita san ever could and also sing in acceptably arbitrary languages. (Among the big stars, Miku v2 / v3 has the poorest solfege skill and strongest tinnitus effect problem, so it would make sense to perfect her first via V5 application.)

On the other hand, V5's CFD computing resources demand would be too high and too expensive as of now, like 8-core Xeon, quad CUDA, etc. This was outright impossible to meet by consumer-customers in the V1 era, but by the time Vocaloid5 engine appears, may be affordable for enthusiasts.