Board Thread:General Discussion/@comment-53539-20130405164209/@comment-53539-20150513183826

@arc, actually Meiko and Kaito didn't loose any range at all. From what I remember, the V1 vocals had only something like a single octave ranges worth of data. What made them seem like they had more was how much "give" the engine had.

Its pure myth they have smaller ranges then the V1 and I'm disappointed people still quote it. In actual fact, they have larger ranges now then they did then. However, the V3 engine had more realism thanks to the way it worked, this added a lot more restrictions then many realize because you can hear more clearly when Kaito or Meiko screw up. So the "give" in the voices no longer exists.

Also, the data gathered for V1 was meant to be analytic centre, which didn't produce as accurate results as one which put samples at the centre of the building of the vocal. Later updates did make V1 use the data as samples and not analytics, the result was it worked more like V2 and beyond, but the results were still quite crappy as V1 still had to improvise a lot. Those who used the last versio of V1 and the original version would have seen quite a result difference.