We need to start talking about the future of UTAU's engine development before it's too late

Tenkoh

As lazy as can be
Defender of Defoko
@Khento does current resamplers and plugins will work on your version ?
Maybe and no. The goal is to make it work with all current resamplers, but since I'm still working on the frontend, I can't say for sure if all of them will be compatible. The idea is to run them under wine on a Linux server. I did some testing and was able to at least invoke m4 and the default resampler remotely, but I couldn't get moresampler to work. Fortunately, Kanru Hua pointed out why, so I'm sure I can find a workaround.

As for plugins, it's impossible to keep compatibility, since they are just Windows executables that get called by the main program. It is possible to implement their features directly on the front-end, though.

Wait, but if this is just a front end gui for it in browser, won't it also become unsupported?
It doesn't rely on UTAU to work, so VB6 support is not a problem.
 

✧ Elfrida ✧

The Space Witch
Defender of Defoko
This is both a really scary and really exciting idea. Mainly because I don't have the most stable internet access right now. If you ever finish this and it goes well I'd highly recommend making a desktop version as well, as its very convenient for a lot of us to not have to halt our progress when our internet goes out. Either way, this looks like a good idea.
 

Tenkoh

As lazy as can be
Defender of Defoko
Working on this is really really fun!
Still working on the front-end, and I've made some progress.

To keep compatibility with UTAU's file format, the track stored internally is always "UST compliant". That means that the front-end will automatically add, remove, resize and merge invisible rests whenever the user adds or modifies a note. In other words, no more rests!
I still haven't finished horizontal movement and resizing, but since the last post I finished vertical movement, lyric editing and horizontal zoom. The code to support different time signatures is already there, and switching between tracks seems to be working now.

Double clicking on a note opens this rudimentary note editing dialog.
kz4ixc8.png


The grid can be zoomed in and out
RZhzFZg.png


When drawing a note on a rest region, the rest will be split and a note will be added in between.
zDZPwQo.png

5H0wNcG.png


Individual notes and regions can be moved vertically.
qY0SuLh.png


GapaDm5.png

I'm probably going to work on a feature to export rudimentary USTs before I tackle the backend. That would make it slightly useful as-is, I guess.
 
Last edited:

Nohkara

Pronouns: He/him
Supporter
Defender of Defoko
Working on this is really really fun!
Still working on the front-end, and I've made some progress.

To keep compatibility with UTAU's file format, the track stored internally is always "UST compliant". That means that the front-end will automatically add, remove, resize and merge invisible rests whenever the user adds or modifies a note. In other words, no more rests!
I still haven't finished horizontal movement and resizing, but since the last post I finished vertical movement, lyric editing and horizontal zoom. The code to support different time signatures is already there, and switching between tracks seems to be working now.

Double clicking on a note opens this rudimentary note editing dialog.
kz4ixc8.png


The grid can be zoomed in and out
RZhzFZg.png


When drawing a note on a rest region, the rest will be split and a note will be added in between.
zDZPwQo.png

5H0wNcG.png


Individual notes and regions can be moved vertically.
qY0SuLh.png


GapaDm5.png

I'm probably going to work on a feature to export rudimentary USTs before I tackle the backend. That would make it slightly useful as-is, I guess.
Please, make a separate thread of this project! I like the depelopment of this project so far~
 

FeatheredFinch

Ritsu's Renegades
Defender of Defoko
Working on this is really really fun!
Still working on the front-end, and I've made some progress.

To keep compatibility with UTAU's file format, the track stored internally is always "UST compliant". That means that the front-end will automatically add, remove, resize and merge invisible rests whenever the user adds or modifies a note. In other words, no more rests!
I still haven't finished horizontal movement and resizing, but since the last post I finished vertical movement, lyric editing and horizontal zoom. The code to support different time signatures is already there, and switching between tracks seems to be working now.

Double clicking on a note opens this rudimentary note editing dialog.
kz4ixc8.png


The grid can be zoomed in and out
RZhzFZg.png


When drawing a note on a rest region, the rest will be split and a note will be added in between.
zDZPwQo.png

5H0wNcG.png


Individual notes and regions can be moved vertically.
qY0SuLh.png


GapaDm5.png

I'm probably going to work on a feature to export rudimentary USTs before I tackle the backend. That would make it slightly useful as-is, I guess.
You should totally request a utalab thread!
 

Tenkoh

As lazy as can be
Defender of Defoko
I'm really interested in following the progress of this project, but I think it would be easier to do so in a separate thread.
Please, make a separate thread of this project! I like the depelopment of this project so far~
You should totally request a utalab thread!

Discussion thread created! http://utaforum.net/threads/melodia-editor-utau-in-a-browser.15867/
I'll post updates there from now on. You can see it live at http://redelee.ddns.net:25565/
 

Arissa

Ritsu's Renegades
Defender of Defoko
Working on this is really really fun!
Still working on the front-end, and I've made some progress.

To keep compatibility with UTAU's file format, the track stored internally is always "UST compliant". That means that the front-end will automatically add, remove, resize and merge invisible rests whenever the user adds or modifies a note. In other words, no more rests!
I still haven't finished horizontal movement and resizing, but since the last post I finished vertical movement, lyric editing and horizontal zoom. The code to support different time signatures is already there, and switching between tracks seems to be working now.

Double clicking on a note opens this rudimentary note editing dialog.
kz4ixc8.png


The grid can be zoomed in and out
RZhzFZg.png


When drawing a note on a rest region, the rest will be split and a note will be added in between.
zDZPwQo.png

5H0wNcG.png


Individual notes and regions can be moved vertically.
qY0SuLh.png


GapaDm5.png

I'm probably going to work on a feature to export rudimentary USTs before I tackle the backend. That would make it slightly useful as-is, I guess.

Dayum, the progress is reallll
 
  • Like
Reactions: SOHBlue and Nohkara

Alessandra

Ruko's Ruffians
Defender of Defoko
Before I begin talking about this, I don't want to make people think that I'm trying to fear or panic-monger; it's just that with every passing year I find this to be a more and more pressing issue that's gotten surprisingly little coverage despite how serious of an issue it could potentially become. So I'm making a thread in the hopes of maybe just getting more people to think about it.

UTAU is free (technically shareware, but functionally free). However, it is not open-source. In a nutshell, this means that the only person who can make any updates to the program whatsoever is Ameya, and Ameya only. This is normally not a problem when a developer is active and regularly pushing updates, and whoever makes whatever software is perfectly free to do with it what they like. But for whatever reason (personal life being busy, unwillingness, whatever), development on the program UTAU has not updated since 2012 (Windows)/2013 (Mac). Hell, the page for Windows UTAU still uses Windows XP as its base recommendation.

In light of recent discussion I've been reading on the topic of emulation and a particularly thought-provoking xkcd comic, the fact of the matter is that UTAU working in Windows 10 completely unscathed was by sheer luck, and we can't count on this luck to continue in future builds of Windows or the Mac OS. Of course, I certainly don't expect UTAU as a phenomenon to continue persisting into all of eternity (especially since its sister program Vocaloid is a closed-source commercial program that's utterly reliant on Yamaha's support), but a closed-source software without an active developer is a ticking time bomb that you can't even read the counter on. Emulators exist, but they decrease in efficiency the deeper you go, and you can't guarantee all the time that they'll work; it would be a shame for a community as thriving and self-sustaining as UTAU to die off because it fell victim to a closed-source software being unsupported.

In addition to the issue of survival, there are a few minor issues as well - UTAU, being primarily self-sustaining, simply evolves far too quickly for a single-manned closed-source program to handle. Right now, to use UTAU to the potential that the community demands of it, you need a giant handful of third-party plugins with incomplete integration with the editor; even common things like VCV and non-Japanese languages require some very hacky solutions because we're bounded by the limits of a program that hasn't seen updates since 2012. On top of that, the differences between the Windows and Mac versions of UTAU caused a minor fracture in the community, which could have been easily mended by developers with both systems working together to resolve those differences. So the open-source model is probably what would work best for the community.

There are a few things in our favor.
  • The nature of UTAU means that a number of its components are already exposed - much of the work is handled by the wavtool and resampler, which can be provided by third parties by default, and so the most pressing issue is making a shell that can contain those functions, oto.ini/prefixmap configuration, and UST files. (This is already a tough task as it is, but it's nowhere near the difficulty of building an inbuilt engine from scratch.) It also means that you can tailor a theoretical new program to fit exactly the comfort needs of existing UTAU users without forcing them into a new learning curve.
  • Secondly, UTAU in its current form is a pretty basic, lightweight program, meaning that the Windows version at least will be likely to survive at least a few more OS migrations. (I cannot be as sure about the Mac version, but that's only because I've never used it much.)
There have been a few attempts at making an open-source equivalent or clone of UTAU, such as Cadencii and OpenUtau, but neither have really gotten off the ground due to the lack of urgency (OpenUtau has had only two developers and hasn't seen a new commit in the last year). The unfortunate reality of the situation is that UTAU is a fairly niche community to begin with, and although we do have talented programmers who have been able to provide us with many wonderful resamplers and plugins, the vast majority of us mainly work as end-users. (I myself am studying computer science in university, but I am nowhere near the skill level required to code something like this from scratch.) That said, I think this is an issue that's been rather low-key for something that should be treated with a lot more urgency, and we need to start thinking about how we want to approach it in the future.

tl;dr: we need to think about how we're going to continue sustaining UTAU in the future or we're gonna die a lot earlier than we should
A frined of mine told me someone is going to release an open source UTAU relaunch... and it's really strange Ameya/Ayame (or ok..) doesn't repl. Maybe now he/she is focusing on other stuff... or just doesn't speak English
 

AykadUTAU

Momo's Minion
No offense to anybody but having UTAU as an online app in a browser is not the best thing to replace UTAU. We need to have a good lightweight actual desktop app that does not require a constant internet connection. When I am on vacation, my parents always choose a holiday house without internet, so if I do not have internet data I will not be able to use it, that is just an example.
In addition, browser apps run slow in some computers. Meanwhile, if we have a desktop app, we will not experience as many bugs...
Also, we need a singing synthesizer that is professional like VOCALOID5, not as complex but close. We also need to have the capability of recording your own voicebank, not using HMM since it sounds strange and not that good. We also need a modern clean-looking UI that is easy to use. We need to keep plugins but not current resamplers, UTAU resamplers sound really buzzy and metallic on low notes compared to some other synthesizers.
 

Mooncat

Teto's Territory
I've been investigating OpenUtau (https://github.com/stakira/OpenUtau) which looks interesting, but I can't get it to work on my computer. As luck would have it, I'm also a software developer, so I could probably fix it. The author, Sugita Akira, has done a great job getting the project where it is, but the bugs hide the fact there's a lot of really good stuff there.

I'm not sure how much progress we can expect from UTAU, and it doesn't help that it's closed-source. If it was open source, we might have a chance to evolve it into something a bit more modern and easy to use. OpenUtau's source code is freely available so that it won't be lost and other people can add to it. It's Windows only, but I imagine the limiting factor is the resamplers, which as (as far as I can tell) Windows native code. Furthermore, the plugins probably wouldn't work, but I'd hope that a new app would negate the need for an awful lot of them.

I'd be keen to hear other people's opinions on OpenUtau and whether it's worth investing considerable time in the project.
 

Kiyoteru

UtaForum power user
Supporter
Defender of Defoko
I've been investigating OpenUtau (https://github.com/stakira/OpenUtau) which looks interesting, but I can't get it to work on my computer. As luck would have it, I'm also a software developer, so I could probably fix it. The author, Sugita Akira, has done a great job getting the project where it is, but the bugs hide the fact there's a lot of really good stuff there.

I'm not sure how much progress we can expect from UTAU, and it doesn't help that it's closed-source. If it was open source, we might have a chance to evolve it into something a bit more modern and easy to use. OpenUtau's source code is freely available so that it won't be lost and other people can add to it. It's Windows only, but I imagine the limiting factor is the resamplers, which as (as far as I can tell) Windows native code. Furthermore, the plugins probably wouldn't work, but I'd hope that a new app would negate the need for an awful lot of them.

I'd be keen to hear other people's opinions on OpenUtau and whether it's worth investing considerable time in the project.
I'm more interested in utsu, personally: https://github.com/titinko/utsu
 
Similar threads
Thread starter Title Forum Replies Date
ProjectKoe Anyway i need help with Oremo UtaHelp 1
thehatman Need help with Teto's rikimi vb UtaHelp 2
Kayuri ⭐️ I need help! UtaHelp 1
S3MiLiA Does my voicebank need glottal stops UtaHelp 2

Similar threads