Mario.MD's NOTEBOOK

May. 11, 2012 at 6:28am with 205 notes

Reblogged from heroofbagels

heroofbagels:

It came to my attention that new UTAU users really need this.

The above visual diagram is simply how I do it when I’m rushing and just need to get an alpha done for testing.
The real meat of this guide is in the descriptions below, which explains why I do it like this.

HOW TO OTO (in five minutes by me)

Firstly, make sure you know how to edit a UST for your UTAUloid in the most basic sense (fitting the UST to your oto settings)
That’s covered here: http://utaforum.net/index.php?topic=393.0

A note; Preutterance (red line) is always placed at the exact point where the consonant transitions into the vowel. This is to prevent offtiming. If it’s in the consonant, the sound will start late. If it’s too far into the vowel, it will come in too early.

The end offset (blue bars) cuts off any fade off you might have on the vowel, and any extra silence. The fade off can be used to effect, depending on voice-type, but in general it should be cut.

Anything covered by the pink bar doesn’t get stretched on long notes. So it covers consonants, and any volume or pitch fluctuations after that. You can use your own judgement on how far into the vowel it goes, but remember; the less white there is to stretch, the more metallic it will sound.

Hard consonant: The overlap (green line) is in the minus area for all of these (eg, -30, -50). The reasoning behind this is that when you sing or speak hard consonants, a small natural gap is left in front of them. Setting overlap to minus replicates this. Front offset simply covers the silence before the consonant.

Soft consonant: Overlap halfway through the consonant. Front offset should cover any silence, and can be used to cover any extra consonant. It’s hard to make some of these sounds short enough to work well, but since you naturally hold them you can just cut them shorter.

Semi-hard consonant: Overlap just before the main part of the consonant. This is a little hard to see depending on the person, and you may have to mess around a little to find the exact spot. If you’re really having trouble, press the “S” in the oto editor to shed some light on the recording. Front offset cuts off silence, but leave just a little room before the consonant.

Vowel: Overlap and preutterance can either go in the same spot around 12-30ms, or just leave them at 0 if you know how to use crossfade. Front offset covers any fade in and off-pitch sections in the beginning of the recording.

This is just a guide and by no means the be all and end all of good otos. If something sounds off, feel free to mess with it! Go crazy and please, figure out by yourself exactly what Overlap and Preutterance do. You can always put them back in place when you mess them up! I feel a hands on experience is required to understand those fully.

I feel like I’m going to get ripped to shreds by Tumblr for this. If you have a problem I’d appreciate if you brought it up with me privately and we can work it out!