Board Thread:Off-Topic Discussion/@comment-25147000-20150320032731/@comment-53539-20150422125318

I think CeVIO has a lot of the issues that UTAU has with sounding like samples strung together, so produces somewhat that familair "autotuned samples" issue I mentioned earlier. Out of CeVIO, UTAU and Vocaloid, Vocaloid has the effect the least while CeVIO the most. However, I'm not going to judge the spoken capablities because spoken word tends to be a lot easier to produce in Japanese. This kinda brings it down to "nit-pick" levels.

In the case of UTAU, I think its because of the fact UTAU is built on a core that was meant for more talking originally and not for singing, and UTAU adapts the core for its use. So this is why it feels so auto-tuned I guess. I know its not 100% like that anymore, but its basically what it was to begin with. :-/

I honestly don't know what core powers CeVIO to explain why it happens. I'm guessing though, the same core powers both the singing and speaking voices, which may explain why it also has that "auto-tuned samples" feeling. I've not really thought much with CeVIO yet because I'm kinda waiting for a voice to really grabbed my attention. UTAU has vocals which interest me and I'm watching to see if UTAU ever changes my mind on it and vocaloid has vocals I like.

What seems to be the biggest shame is how some voices in CeVIO either use one or the other voicebank method and few have both singing and speaking capabilities.

I'm not sure how I feel on "One". Though I feel IA produces much better overall results.