Utaus can sing better than Vocaloids (3 and 4) in full 2016 ?

[Haruna-Chan]

A simple question today, it has as a UTAU (VCV + CV/CV-YOU Extra CCV-VCC) sound better, equal or similar to Vocaloid (3 and 4)?

I've been using resamplers (as the moresampler 0.6.1) and I realized that every time but mine and utaus that I'm using are (at least for min) sounding better than Vocaloids (3 and 4) or even completely human in almost all languages what I realized an improvement mostruoza was in Japanese, Spanish, English and Portuguese.

 

[★StarsNeverStop★]

Sometimes. It depends on whos working with them and the voicebank quality too. Though i've heard even simple CV UTAU(s) sound human and beautiful just with a little effort by the tuner

 

[keopi]

can we hear some examples? :'D

 

[Haruna-Chan]

can we hear some examples? :'D

Of course!, here a good example (at least for min sounds better than V3)
[PT BR]

 

[na4a4a]

Vocaloids, like Utau banks, vary wildly in quality.
Not only this but there are very few companies producing them so there isn't a lot of space to have amazing vocals...

Also Vocaloid kinda takes a slightly different approach that makes it much smoother sounding...but perhaps a tad more plain and unflattering...or gross...

ofc we have voices like sweet ann and avanna that are decent quality on the english side but very specific in their usage.... imo literally every other english voice is lacking quality-wise.let's be real, they have no standard....we have voices like sonika, big al, and cyber diva to show that.
In terms of Spanish, Maika and Bruno are very very good quality.
Japanese is a crapshoot but there are so many voices that it doesn't matter.

So Utau (or Vocaloid) isn't really better, just that Vocaloid different and has a limited and underwhelming selection or actually good voices.

 

[수연 <Suyeon>]

Vocaloid: quality is a crapshoot. it used to be that "studio quality" was to be expected from every voice, but... we have Sonika, Dex, Daina, and Ruby as proof that this isn't necessarily true across the board. not to mention that English varies wildly between versions (even banks built in the same version all have different reclists). while the result is generally smoother, they can sound bland without editing (or unintelligible in the case of soft voices like Yuki, Luka V2, Zunko, etc).

Utau: quality is russian roulette. anyone with access to 1) a computer, 2) audacity, 3) a mic (regardless if it's their laptop, cellphone, an el cheapo, or XLR set up) can make a bank. because a lot of banks done by people aren't in their native language: accent varies from almost native to foreigner who never took a Japanese class/bad anime dub quality. there's also the varying oto quality, tuning, and editing mastery to widen the gap.

frankly, both can sound either human or like poo depending on the factors that effect the voice (mic and environment, reclist, skill of the voicer, skill of the programmer, tuning vs raw, etc.).