Imagine You Are an UTAU Dev...

WendytheCreeper

(>☉ ͡ヮ☉<)
Defender of Defoko
Imagine that you are an official developer of the UTAU software. Assuming you had unlimited programming ability, time, resources, magic, etc...What would your ideal version of UTAU look like? What features would it have? Workflow? What languages would you try to optimize for? Or would you try to optimize all of them? Would it be standalone? In a DAW? Go nuts! I don't want to just hear your everyday performance enhancements, I want to see how you would imagine UTAU!
 

Buck

Ruko's Ruffians
Supporter
Defender of Defoko
Where do i begin

-Itd probably be standalone, based on how A/E has been, a VST based vocal synth is rather tedious to control.
-It would probably look and work more like a DAW like Caustic or something, where it revolves around creating unique patterns and then putting them together inside a sequencer section.
-I would try to make the language system as neutral as possible. I'd like to think that if you focus on the individual phonetics rather than specific words than language wouldn't be a hugely limiting factor.

I'd make properties editing more graphically inclined, with knobs and reference images instead of number boxes.

Tl;dr, id make it so it looks like it was made by someone who planned to use it a lot.

I should learn to code audio stuff XD
 

KNΞMΛTCS

Just an UtaForum user
Defender of Defoko
First off remove the need for rest notes, add real time rendering, and make it have US locale support proper. Then, give it a piano roll like the one in FL Studio (or any good daw). Have every single parameter editable in automation similar to vocaloid, but in the form of clips you can reuse wherever, and with point editing. Pitch tuning can stay as is, just with the points easier to hit with the mouse. Full multi tracking is a must, as is the ability to import wav/mp3 files. Throw in some quick eq/fx functions that can be tied into the voice, give it a dark theme ui with a buttload of keyboard shortcuts, and there you have it.
 

Agatechlo

Specified.
Supporter
Defender of Defoko
First & foremost, I'd make UTAU 100% Unicode compatible - no more having to switch to Japanese locale to install it (I managed to install/use it without any Japanese language support on this system, but it's something I don't recommend for the common user). Then add full language support for English, Spanish, Portuguese, German, Korean, Chinese & Tagalog; add additional languages per user demand.

I agree with Kinematics on getting rid of rest notes, multitracking & importing mp3 like Vocaloid, but I kind of like how UTAU handles pitch bends, gender factor etc. on a per-note basis. It's really hard to move notes around in Vocaloid because all the parameters have to be moved separately.
 

Iris Hunter

Momo's Minion
  • I'd definitely start with fixing the locale. It's such a hassle.
  • I'd definitely work on the UI too. Not that It's bad,but I think it's begging for improvements.
  • I'd definitely make it work with any language. Let it be Japanese or some language someone just created.
  • I...guess I'd make a plugin of it too...While I would like it, I don't know if the majority of people would like it or not.
  • VSQx support. Just a hassle to save it as VSQ in V3/v4.
 

na4a4a

Outwardly Opinionated and Harshly Critical
Supporter
Defender of Defoko
  • Start from scratch probably because the code is really old
  • remove need for rests
  • No US locale support, just make the whole thing Unicode, probably UTF-16 encoding since it would use less space on non-alphanumeric characters.
  • UI could be changed to be more flat "metro" or "material" design as that's all the rage now. Maybe ability to change color palette form a few options.
    • smoother scaling/resizing of the editor. Including the size of the notes.
  • Proper support for importing of the Midi and Vocaloid pitch parameter.
  • Remove envelopes, keep it internal and largely hidden unless you're making a vb and need to test, make everything pertaining to volume a continuous curve.
  • Oto and all that can stay 100% the same because there is nothing actually wrong with it.
  • Configuration limit should be extended
  • Dictionary support where you can tie a word to specific alias groups and timings, this would be per-voicebank.
  • Realtime rendering for low quality preview and them high quality but slower exporting.
    • or if the system is fast enough, use of a buffer where the high quality version is rendered as it's played.
  • VST/rewire bridge that lets you use midi tracks in your DAW in Utau. But Utau still remains standalone.
  • True multitrack support.
    • Use of different voicebanks per-track
  • Ability to call multiple voicebanks (think appends) and call them with suffixes while still keeping said voices separate.
  • Easy interface translation with a single file.
  • Replace file format with something like a modified midi format or MusicXML.
  • Alternate staff notation view.
  • Ability to play a backing track along the project.
  • Separate the voice creator from the project editor but allow from the creator to hook into the editor if both are available.
  • Fix or replace the disaster that is the default resampler
  • get rid of unnecessary flags as they often confuse people and are used blindly and wildly.
  • contain banks withing a write-protected file so that vb makers can prevent editing (maybe make this a paid/donation feature)
  • Allow you to specify if a sample is ending in a consonant so that the creator can just gate the noise (if any) after it.
  • S-curve fade-out on the final note of a phrase so it's more natural without the need for end breaths.
  • End breaths can be automatic and transparently added with a "-" at the end of any sound, like a suffix.
  • brush it's teeth
  • new makeup
  • walk walk fashion baby.
 
Similar threads

Similar threads