Latest update Musescore 2.1 on Ubuntu

• Feb 26, 2018 - 14:10

I received this useless and unnecessary update maybe five or so days ago and there are changes in it that are simply retrogressive. I don't know who is in charge of the Fluid soundfont, but I really wish they would just leave things alone. The orchestral percussion has been fiddled with and the options for any of the cymbals and snare drums used in an orchestral score are useless. All of the snares sound like a tinny machine gun during rolls. This was not the case previously. The only orchestral cymbals that sound right are ridiculously quiet, the complete opposite of what these cymbals are supposed to be. The alteration of the soundfont has completely messed-up the instrument settings of every score I have saved.
Playback now judders like crazy. This was never a problem even when running numerous sfz soundfiles at the same time, which are more resources hungry.

I don't know why things that aren't broken can't just be left alone.


Comments

There are some known regression in 2.1, esp. in the area you describe. 2.2 will have these sorted.
Check https://musescore.org/en/handbook/developers-handbook/release-notes/wip… and https://musescore.org/en/handbook/musescore-22-hit-list.
I guess you are talking about #196321: Tenor Drum playback not functioning correctly (see also https://musescore.org/en/handbook/release-notes-musescore-21#known-inco…, item 7) and #197131: Regression: Bass synthesizer and bass guitar create popping sounds. The latter is fixed in 2.2, the former will be shortly before release
If it is just the soundfont, you can grab the 2.0.3 version of it and use that with 2.1

Well, some of these problems were reported, but a number sound new to me. If you want to make sure they are fixed in 2.2 - which is coming out soon, 2.1 was released many months ago - please attach your score and describe the specific problems in more detail so we can investigate.

Overall, 2.1 contains literally hundreds of major improvements over the previous release, so it definitely isn't useless. But indeed, as with any new release, occasionally while fixing hundreds of bugs and implementing dozens of very useful new features, a handful of things might break. Such is the nature of software development, and why we welcome feedback from users like you reporting the problems you discover so we can fix them!

In reply to by Marc Sabatella

That's really your standard reply to 99% of the questions on this forum: "attach your score". It might astonish you to know other people are also extremely familiar with using musescore, and its structure, and don't need "score doctors" to work out problems that are part of the software's release cycle.
I do builds from source code on my other two computers, but thought maybe I would rely on software updates for my Ubuntu machine in the way that any general user would. That is the crucial point: most will just be downloading the software as a package or an update and these really ought to be fully tested beforehand. These are simple, non-code issues like poor instrument sounds that don't need multiple reports.
So I reject your "attach your score" request because it is incoherent with regard to this issue. They originate in the builds and changes. Please try and remember this.

In reply to by Jojo-Schmitz

Not a really enlightened riposte. How does this address the issue that an official release is riddled with problems that could and should have been tested beforehand. That is before releasing to average users depending upon the software working?
In what way was it impossible to to know that the soundfont is inferior to its predecessor? It only requires normally functioning hearing.

It isn't my job to fix bugs. I have not attached myself as a developer. But if I had I doubt I would be telling other people to fix them after having caused the cock-up myself.

In reply to by Roger v.d Velde

You reports a bug, you don't want to fix it, so you need to help those that are willing to fix it, simple as that. You could have testes 2.1 more than a year ago, there was an RC prior to release. And even before that numerous users helped in test it.
If you want to report a bug, by all means do so, but please explain exactly what that bug is, what you do, what you expect and what you get instead, and how to reproduce.
If you don't want to help, fine, but then please stop complaining.

The default soundfont did not change, not significantly, nothing that explains why you believe to sense a decrease in quality, it did change some drumset assignment and did fix a few pitches for some instruments that had bad samples or where out of tune.
There was indeed a regression reg. Tenor drums, and about some bass synthesizers, the latter not related to the soundfont itself at all, the former has a valid workaround of using an adjusted drm setting.
ANother sound related problem had been fixed the other day, again not related to the soundfont itself, but to fluid, the synthesizer, yet another is still in the works (but very unlikely yours)
I'm not aware of any changes to cymbals or snares.

In reply to by Roger v.d Velde

As mentioned, the soundfont barely changed at all, and the few changes that were made were mostly improvements to specific notes on specific instruments that were out of tune or whatever. The changes you are perceiving are almost certainly more due to the code than the soundfont, and those changes did fix problems, so that for very many scores, the sound is improved. But indeed, there are some small subset of scores that sound worse because they use particular instruments in particular ways that these code changes did not handle well. So for that small subset of scores, indeed, the perception might be that the sound has gotten worse. It's unfortunate that at no time during the extensive test phase for 2.1 did someone who had a score that would trigger these problems try it out for themselves, because if they had and had reported the problem, it could have been fixed. But as it is, the very many people who volunteer their time to work on these releases had only their own scores to test with, not yours, and so no one was able to detect the problem.

The lesson here is, open source software development really is a community effort, and it depends on the community of users to help test as much as it depends on the community of programmers to do the development. That's the way open source software works. We would welcome your participation in the 2.2 development process, I strongly encourage you to try a build of 2.2 to see if the fixes already made solve all your problems or not. If they don't, again, please help us help you by sharing a score that demonstrates a problem.

In reply to by Roger v.d Velde

Well, yes, in order for us to understand the precise problem you are encountering, we need to be able to reproduce the problem, and in order to reproduce it, we generally need to see the score. Or, alternatively, have precise step-by-step instructions to reproduce the problem from scratch.

Testing is a complicated thing. MuseScore is open source, meaning people from all over the world contribute to its development as well as its testing. Pre-release builds of 2.1 were available for months before its release, and many people helped test it. In the process, dozens of other regressions were found and fixed, but its pretty much inevitable with software development that no matter how much you test, something will fall through the cracks, and this is what happened here.

The changes to the sounds in 2.1 were in part due to changes in the samples themselves, but more likely the changes you are concerned with came from changes to the parts of the code that handles the playback. These code changes were made to fixed other problems that had been reported in the past and worked quite well for that purpose, but unfortunately these changes introduced new problems that went undetected until too late because none of the open source community of people testing the software happened to encoutner those cases. Again, we can't fix problems that don't get detected and reported.

Anyhow, if you want us to fix the issues you are encountering, we will need to know how to reproduce the problems. If you reject the request to attach your score, there unfortunately isn't much we can do. The problems that were reported have either already been fixed in the code or are actively being worked on. But ones that don't get reported in a way that allows us to reproduce the problem can't be fixed. You can either hope the problems you are concerned about are among those that other people have been willing to help us reproduce, or you can take out a free insurance policy by helping us reproduce the problem yourself.

In reply to by Marc Sabatella

I really don't want to see or hear that request about attaching scores because it's completely irrelevant. I think I already made that clear. Seeing that my score is linked to an orchestral drumset is no guide to how it sounds, or that the sound has changed. I really can't fathom how this simple fact is eluding you.

Examining a CD collection doesn't fix a fault in a CD player.

In reply to by Roger v.d Velde

I can only assure youwant to see this problem fixed. I assume you do as well. So all I can do is repeat myself - if you want to see a problem fixed, we need to be able to reproduce it. Otherwise, we probably can't fix it. A verbal description of the problem can be enough to give us an idea of where to look, and since others have posted scores, we've been able to fix several problems. Maybe those are the same as the ones you are seeing, but the only way to be sure we've fixed each of the specific problems you are seeing is to have a score or scores that reproduces the specific problem.

To borrow your analogy: if a CD player plays 1,000,000 CD's with no difficulty, but six users say they have a problem, that does suggest one or more faults in the CD player indeed. But the CD manufacturer probably won't be able to identify or fix any of those faults without a copy of the CD that triggers the problem. If five of those six users choose to share CD's that trigger the problems they are seeing, and the CD manufacturer is able to identify and fix three problems that turn out to the ones responsible for all of the issues reported by those five users, great. But there still be no way of knowing if the problems experienced by that sixth user - the one who refused to share his CD - were all covered by those three fixes, or if one of the problems he is experiencing is perhaps unique to his score. Either the fixes that the manufacturer makes address his problems or they don't - no way to tell unless he shares his CD.

Up to you on whether you want to take the chance that the fixes already made fix your problems or not. If they do, you will be happy when 2.2 comes out and your problems are over. But if not, you'll have only yourself to blame.

I was going to say this is like that thing people will tell you where if you don't vote in an election, you have no right to complain about the result. But actually, the situation here is more direct. With an election, if you vote, the election might still go against your wish - there is no guarantee that participating in the process will actually lead to the result you want. It's really barely better than lottery odds of it. In the case at hand, though, there is excellent chance that helping us out here by sharing a score will allow us to verify if the problems you personally are experiencing are fixed already or not - and if not, your sharing a score will be exactly what makes the difference in our being able to fix it. it's a no-lose, huge-win scenario.

Your choice.

In reply to by Marc Sabatella

The scores merely point to functions of the particular software. The scores do everything they need to do. If I copy them to another machine containing the previous version of Musescore they work fine, but after that update the soundfont the score pointed to had changed. It's really that simple. Having a copy of my scores with their instrumentation altered to fit my particular extra instruments would be of no help to you or anyone else. To see what I described would require you to visit my house and see it on my particular machine. The irritating pretence at scientific method is tiring. As is the veiled suggestion that I'm somehow making elementary mistakes you will be able to rectify. That is not the case.
I merely am pointing out that this update has faults in it that some people will just accept as is an update and that some parts of the soundfont are altered, among other things. Something that was acknowledged in the first reply. Take it or leave it. It's a failure of software production.

Absolutely exasperating.

In reply to by Roger v.d Velde

I am sorry I have been unsuccessful in my attempts to explain why we would need to see a score in order to reproduce the problem you are hearing and be sure it is fixed for the next release. The miscommunication is indeed exasperating to me as well.

Anyhow, I do truly hope that the fixes already made happen to address the problems. I would love to be able to test that and be 100% sure, but without a score, all either of can do is cross my fingers and hope for the best.

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