General questions about UTAU/OTO..

Seiro

Teto's Territory
Sorry for my frequent help posts, but I have more questions...

1. How do I get an UTAU to sound less choppy near the end of notes? I'm using Teto's english VB, and on her last note she stops abruptly instead of something like a fade out. Can I do something to achieve a fade out effect, or is that usually done in DAW's?

2. I'm recording a VCCV VB right now, but I notice on the parts where I am not speaking, there's a bit of white noise. I can get rid of it in audacity after I record so it sounds crystal clear, but would that be necessary in the first place?

Thanks!!
 

Kiyoteru

UtaForum power user
Supporter
Defender of Defoko
1. How do I get an UTAU to sound less choppy near the end of notes? I'm using Teto's english VB, and on her last note she stops abruptly instead of something like a fade out. Can I do something to achieve a fade out effect, or is that usually done in DAW's?

It depends on whether the note ends in a consonant or a vowel. If it ends in a vowel then yes, you can certainly edit the envelopes to have it fade out gradually. If it ends in a consonant, be sure to use all the transitional samples that you have available. Additionally, you may find that you're actually just ending the note sooner than expected with the consonants, and the sudden stop is making it seem choppy, so consider lengthening the note a bit.

2. I'm recording a VCCV VB right now, but I notice on the parts where I am not speaking, there's a bit of white noise. I can get rid of it in audacity after I record so it sounds crystal clear, but would that be necessary in the first place?

Please SING your samples rather than speak them. Especially with English voicebanks, there's a noticeable difference in tone and pronunciation between samples that are sung and spoken. By sustaining notes, you will have a looser and more musical sound.
Regarding the noise, please leave it untouched. I would not advise cleaning or editing samples unless you have robust software for that purpose. Audacity's noise removal often removes important harmonics from the original vocal and leaves it sounding worse than before.
 

Similar threads