Help with my UTAU Rerecording

Kiyoteru

UtaForum power user
Supporter
Defender of Defoko
Copied and pasted from earlier PM.

Some of your samples start very close to the beginning of the file, so there's not enough silent space and things get cramped. Try to have a bit of quiet at the start of your samples, since the OTO can handle cutting it anyway.
The Rs seem to be English Rs. A Japanese R is closest to an alveolar tap, and you can probably look up videos about how to pronounce it. Alternatively, you can simply record them as Ls.
Some of the Os sound a bit more like "oh" than a plain "o." It seems like you've controlled it in other samples, so it's just a matter of checking your recordings through the next time you make a voicebank.

Anyway, just to make sure that you have all the CV samples you need, you can use oremo's default CV reclist and stop once you get to "pyo". Here's a romaji version of it. https://app.box.com/s/o7ew28dzyt2f1gey1l9w

Also, might I suggest doing a CVVC bank instead of VCV as your next step? It's less recording and less otoing, while being as smooth as VCV, and not needing any UST editing beyond maybe converting from romaji to hiragana if you have presamp set up.
 
  • Like
Reactions: Zoku

Similar threads