Soundfont & MuseScore 4

• Nov 14, 2022 - 19:52

Is there any way at the moment to add a soundfont to MS 4 so my Theremin is recognised and given the correct sound, even if it is only a temporary solution?


Sure, just specify the soundfont in the mixer. Since Theremin isn a General MIDI instruments, MuseScore won be able to pick the sound out of a soundfont containing many sounds, so just be sure itś the first (or only) sound in the soundfont. You can also use a VST instrument which is likely produce better results, if you can find one.

In reply to by Philorganic

Hard to say from just a picture - much better if you attach your actual score. But it does look like for some reason if you add theremin, you don't get a mixer channel. You should report that as a bug on GitHub. Meanwhile, just add some other instrument - you're going to change the sound anyhow, and then also change the name.

In reply to by jeetee

Not on Linux.

This cabability was done so well in version 3.6 that I used it heavily. And the reverb.

Am I complaining? Yes. Hopefully soon this will be fixed, because if not, I am eventually NOT going to be able to use Musescore 3.6.2 on Linux, and I will be very unhappy.

In reply to by Chuck Bermingham

As mentioned this feature certainly returning, and in a far better form. Meanwhile, you can simply split any non-GM soundfonts into separate files; a number of utilities exist for this purpose. Then you can use them directly in MU4 as easily as in MU3. It's also quite reasonable to expect that at some point, a Linux user with the necessary motivation and skills will implement VST support, although the limiter there is that there are very few VST's available for Linux to begin with, which is probably a big part of why so far no one has volunteered to do this.

Meanwhile, 4.1 is due out this week and contains a new built-in reverb. In some ways it's much more configurable than MU3 - completely independent control for each channel . In others less (no controls over the various internal parameters that are then globally applied to all channels). But for that sort of global sledgehammer-style reverb applied to the already mixed output, it's trivial to add in Audacity or whatever.

In reply to by Marc Sabatella

What does "far better form" mean, and when can we expect it? It's not in the 4.1 beta.

I've been advised, time and time again, I won't have VST support in Linux. And again, I was so happy with what I could do with 3.6.2, on my own platform, and it's starting to sound as if I won't be able to do it anymore, once the distros stop providing the necessary libraries to keep 3.6.2 running.

Very disappointed. Took a crack at doing another Buxtehude organ piece on 4.1, and it was pretty pathetic.

As regards the reverb: sure, I can send the .wav file through Audacious, or even Sox, but one of the most important reasons I started using Musescore again rather than fighting with with LMMS is because Musescore got the Zita reverb, and as any organist knows, being able to work with live stop changes in a church helps make things sound as they should, working note by note. Can't do that in Audacious.

In reply to by Chuck Bermingham

"Far better form" means it is planned to organize the list better, provide search facilities, etc. In MO3, having multiple soundfonts loaded meant having to scroll through one interminably long list try to find the specific sound you're looking for. This was a problem people reported over and over. And that's one reason simply reimplenebnting the awkward existing control was not seriously considered, I think - they want to make things better.

As for VST in Linux, not sure what you mean. No one has ever said it won't happen ever. Just that it hasn't happened yet for reasons that have been pretty carefully explained. It could have happened already with support from the core team had someone volunteered to start the project, but apparently there just isn't that much interest since no one has yet volunteered. I'm sure that will change at some point.

Can you explain what in particular you had trouble with in 4.1, though? As I said, reverb is there, and if soudnfonts are split into files, you can continue to work with them in one interminably long list as before. So there should be little difference, other than the inability to tweak internal parameters of the global reverb. But you refer to do something note by note; that most certainly was never possible in MU3. That would require an external DAW. So I'm confused. If you start a new thread and attach on of your MU3 scores we can understand and assist better in upgrading it to take advantage of the phenomenal engraving improvements in MU4 while not giving up anything - and perhaps gaining something - in playback.

In reply to by Marc Sabatella

I'll wait until version 4 is actually ready, and continue using 3.6.2 until then.

Many years ago, I was dragged into a database project, because the former project leader ws telling his users to use whiteout to address the deficiencies in the completely ridiculous output he was generating. took me over 3 months and a consultant to fix his mess.

This is starting to look like that situation. I hope to Heaven that changes.

Split soundfonts into individual files. Inadequate reverb. Yikes.

In reply to by Chuck Bermingham

Not sure why you are thinking separates file for separate sounds is a terrible thing. It's the way every single format except sf2 works. Far more efficient.

|The reverb, as mentioned, is actually better in virtually every single way compared to MU3. But it could be that the one limitation is has might turn out to be more important for your specific score than it's advantages; we won't know until you start the new thread thread asking for help with that.

In reply to by Chuck Bermingham

Glad you're finding that useful! I'm not understanding what you mean about instrument changes, though. The usual point of them is to change the sound (as well as perhaps the notation). So indeed, after adding an instrument change, if you want a custom sound for the new instrument as well, you'd want to select the desired sound in the mixer just as you did for the original sound - except of course you'd normally choose a different sound.

In reply to by graffesmusic

Yes, but many of these are just simple effect plugins that MuseScore will be providing native versions of (the reverb in 4.1 is just the first step). None of them are anything like the large high-quality sample libraries available for Windows or macOS. Not saying it doesn't make VST worthwhile on Linux, but it no doubt has been a factor in why no Linux developer seems to have yet been sufficiently motivated to implement it even after almost a year since the original 4.0 beta when they could have begun the work.

In reply to by Marc Sabatella

So, any idea how Musescore 3.6.2 turned out so effective for Linux users? Is there someone who doesn't want to work on the project anymore?

I would be happy to jump in myself, except I have very little time to do anything other than keep up my domestic situation (without going into details,) so having 3.6.2 available was very helpful for me as a way to relax and be creative in the way I wanted to.

I hope the developers aren't making light of my needs and requirements.

I will wait and see.

In reply to by Chuck Bermingham

I can't speak in any official capacity, just as someone who has been involved with the project longer than just about anyone likely to be seeing this. So what am sharing here is my own personal perspective.

The founder and original lead developer Werner Schweer was a Linux user. So MuseScore started out on Linux, where he and a dozen or so other people enjoyed using and developing it there for a few years unbeknownst to the vast majority of musicians. Then some Windows and macOS users volunteered to port it to those systems back in the early 2000's, whereupon the popularity of MsueScore instantly skyrocketed, and most of the subsequent development has come from that community. Werner retired a few years ago, but there are certainly some Linux users still involved - it's just a much smaller community.

MuseScore continues to run incredibly well on Linux, and 99.9% of functionality is identical between all operating systems, because almost all of the source code is system-independent or uses libraries like Qt and FluidSynth that handle the system-dependent stuff well enough that most things just work on all systems with no special effort required. That's why it's basically never been an issue until now.

But the new playback system - including Muse Sounds and VST support - is one of the most system-dependent new pieces of functionality that MuseScore has tried to support since those initial early days of just getting the program running on all three systems to begin with. And there are no libraries like Qt or FluidSynth to do the hard work of supporting VST on all systems. So for the first time in the past 20 years or so, there has had to be a significant amount of system-dependent work done completely from scratch, separately for each OS. It was a monumental undertaking (that I was not directly involved in - I'm not an internals guy at all).

While the release of 4.0 could of course have been delayed while waiting for someone sufficiently expert in Linux system internals to finish getting VST support working, it's hard to make a case that this would have been fair to the MuseScore community as a whole. Since VST was working on Windows and macOS and thus well over 90% of MuseScore users can use it, and since relatively few Linux users rely on VST compared to Windows or macOS since there is so little support for it in the industry as a whole, - and since and Muse Sounds are generally superior and work on all systems, and so do soundfonts, and since global "sledgehammer"-style reverb is trivial to add externally, and since accomplishing tasks by using multiple programs that each do one thing well is absolutely the Linux way - it's really hard to make a case that it would have made any sense at all to delay.

Frankly, though, given this situation has been known for almost a year, I'm still quite surprised no Linux expert has volunteered to address it. But again, I think this speaks to the relative small size of the Linux community, and the relatively small amount of interest in VST within that community.

In reply to by Marc Sabatella

Just keeps getting more and more depressing.

It was so simple: I saw a well-written piece of software that could do what I wanted to do for many years, I started using it, and now what?

3.6.2 still works, and does a great job, but for how long?

I needed 4 things to make this all work:

  • A good notation system (iused o use Quickscore Elite Level II on Windows)

  • A configurable real-time reverb (ditto)

  • A great organ soundfont (stefans.)

  • The ability to write .wav files (I believe it was possible when QSE went to Fluid.)

When MS3 came out, I was elated. Again, now what?

Like I said, worse and worse. I hope to God I'm wrong. Those 4 things should work on all platforms.

In reply to by Chuck Bermingham

You are indeed wrong. The notation system has gotten tons better. MU4 leaves MU3 in the dust here, no comparison whatsoever. Prior to MU4, MuseScore was the worst of the major notation programs. Now it's one of the best. The reverb has, as I described, gotten better in most ways - more configurable than MU3 was with respect to channels . MuseScore continues to support soundfonts. And nothing has changed with respect to WAV. So overall an improvement.

It's still not at all clear what you are finding to be a problem, but this thread has run its course. If you'd like us to help you take advantage of the enormous improvements in MU4, we need you to start a new thread and attach an actual score as I explained earlier. We can't help you further otherwise. But, if you don't value good notation, then you are of course welcome to keep using 3.6.2 as long as you like.

In reply to by Marc Sabatella

Thought about it and yet, I will start a new thread and see what ensues.

There are a few minor problems with the score I will submit, bu they are minor enough that I was still happy to ost it to youtube and facebook as audio.

Still, I will wait and see. I am not comfortable about jumping through all kinds of hoops to get this performance sounding right, and I would be very sad if the Musescore team is going to sacrifice that for "beautiful notation" on my platform.

In reply to by Chuck Bermingham

FWIW. I found a version of Stefans organ sf2. You can load it into the MU4 soundfont file and use it in MJ4. But it is only one sound.

Or you can download and install Sforzando. I wouldn't pretend to know how to use it. So I did it the hard way.

I believe that MU4 will recognize Sforzando player automatically. Open an organ score. In the mixer, select the down arrow on the right side of the top of the organ track, Select VST3/Plogue Art/Sforzando. Hopefully the player opens. Then I dragged the Stefans font onto the player. I had access to several differ3nt sounds which worked in MU4. I have no idea about changing sounds during the piece. I know that is of upmost importance in organ music.

Dynamics sortof work. But only for all staves. There may be a better way to load fonts into Sforzando. I only went this far to see if it could be done. If you use the Hub to download the reverb effect, it works well on the organ.

Bottom line: It is possible to use your organ font in MU4. You would need to spend a little time in Sforzando to learn how to use it. Reverb work well.

Is it worth it? Only you can tell. I think once it is all set up, it wouldn't be any more difficult to use that other things.

In reply to by Chuck Bermingham

I think I'll stick with 3.6.2, and see what happens with 4.x, thank you.

I am aware of the notation enhancements. Those are great, but they don't address my problems.

BTW, one of the things I will be waiting for is the return of the piano roll, or at least a way to adjust the playback length of notes without changing their appearance on the screen.

I am not splitting soundfonts.

In reply to by [DELETED] 20089696

Also, with support for VST and also Muse Sounds, it will be less common to need alternate soundfonts in the first place. And as mentioned, there are VST instruments that will allow you to select sounds within soundfonts. There are also utilities that can extract individual sounds from soundfonts. So pretty much no matter what your specific need is, it can be addressed. If you explain more about what alternate sounds you are using and why, we can help you get the job done.

In reply to by Marc Sabatella

Seems like I'll be covered then. I normally select different instruments from different soundfonts, but if we are going to have per instrument VST this will be just a little more tricky. Another question regarding MS4 —perhaps a little off-topic—, will it have better velocity and volume handling? From what I know, I am only able to change velocity locally (per note) via inspector or piano roll, but that seems not to be the case of volume.

Kind regards and thank y'all for the replies!

In reply to by Omicronrg9

Are you making a distinction between velocity and volume? MIDI doesn't really have a standard way of separating these concepts, although MuseScore 3 does for instruments supporting single note dynamics. In the initial 4.0.0 release, there won't be any controls here, but by 4.1 or 4.2 or so, it's planned to have more sophisticated controls, similar to the "lanes" provided by some DAW software.

Maybe this is in the works, I will look over github.

Soundfonts in MuseScore 4: When this is fully supported, maybe sort the fonts by bank number, then program number?

The numbers need not be visible in the list, but it would definitely help, especially with soundfonts that have banks with lots of related presets.

Do you still have an unanswered question? Please log in first to post your question.