Runtime Error

• Nov 17, 2015 - 21:14

I created a score with odd time signatures. 13/32, 9/32, 4/16, 6/16, all for different instruments in the same instance. Then immediately after the passage that requires those times is over, I set everything back to common time. Whenever I try to listen to the whole piece, an error comes up saying "Runtime Error! Program: C:\Program Files (x86)\MuseScore 2\bin\MuseScore.exe. This application has requested the Runtime to terminate it in an unusual way. Please contact the application's support team for more information."

This only happens when I listen to the part with the irregular time signatures. I can listen to anything before it or after it without a problem. The program crashes after getting this message. How do I fix it ?

Comments

• Nov 18, 2015 - 01:06

Well, unless you are a C++ programmer willing to help with development of MuseScore, you probably can't fix it yourself :-). But you can certainly help *us* fix it. First step would be to attach the score and tell us how to reproduce the problem.

Reply

• Nov 19, 2015 - 22:53

Custom time signaures should not present problems, but there are a couple of known bugs with using multiple time signagures (custom or otherwise) at once - an experimental feature known as "local time signatures". None of the known bugs should result in a crahs or affect playback, but I wouldn't be at all surprised to find out there were other bugs with local time signatures. Like I said before, if you'd like this fixed, we'd need you to post the score you are having problems with and steps to reproduce the problem.

Reply

• Nov 20, 2015 - 02:27

What I don't understand is that you need the score to work on the issue !? It's a bug that a lot of people are having ! Why should I need to provide my hard work to you people who probably won't fix it anyways ? I'm telling you what the problem is. Recreate it yourself ! Or at least give me a good reason why I need to link my work. This is ridiculous. You should know how to fix these bugs on a dime and give us patches/solutions/techniques. And what kind of score program doesn't support local, custom time signatures varying among instruments without completely screwing up the score ?!

Reply

• Nov 20, 2015 - 02:53

It's fairly straightforward procedure—certainly "recreate it yourself" is an essential step before fixing the bug, but it would enable that much more quickly if you could describe how to reproduce it. If you want to keep your music private, you can fill in random notes or something before attaching the score here. It would certainly be worth the effort if the next version of MuseScore lets you do this kind of work bug-free, don't you think?

Reply

• Nov 20, 2015 - 03:45

I read forums thoroughly every single day and have for the past five years, and I don't recall anyone ever posting a problem like yours. Perhaps I missed a report somewhere that you saw; if so, feel free to post a link to that other report.

Not sure what makes you think we won't try to fix it if we can; that's what we do. We fix tons and tons of bugs. But we can't fix bugs that we don't know how to reproduce. I have tested dozens of scores that use custom and local time signatures and never seen a crash such as you are describing. So something must be unique about your score, or the way you went about creating. Thus, we need to see the score if we are to reproduce and fix the problem. Feel free to delete content - irrelevant measure, some or all of the notes themselves, etc - if for some reason you have concerns about sharing your score.

As for what kind of score program does not support different time signatures on different staves, actually *most* programs do not support this. It's an advanced fwature found in only a very few programs. And most of the programs that do support it cost hundreds of dollars. MuseScore is free, developed and supported by volunteers. Please try to be considerate of the enormous effort that is spent providing this software and support.

Reply

• Nov 20, 2015 - 06:24

Fine, here's the score.

The problem starts at 93. The motif I'm having the problem with goes through 97. I can play everything else except this. It gives me the runtime error and that it caused the program to terminate it in an unusual way.

I created another score with this same part, only the trombones with local time sig. and the same error occurred as soon as the playback passed through the bar.

Attachment Size
North American Suite I, Mvmt I.mscz 77.51 KB

Reply

• Nov 20, 2015 - 09:19

There's something seriously wrong with this score - it stops the JACK server after the first note when you try to play it.

There are also some crazy corruptions reported....

Bar 94 Stave 8 incomplete. Expected: 13/32; Found: 11/32
Bar 94 Stave 18 incomplete. Expected: 13/32; Found: 52/2304
Bar 94 Stave 19 incomplete. Expected: 13/32; Found: 13/72
Bar 95 Stave 7 incomplete. Expected: 13/32; Found: 39/2304
Bar 96 Stave 1 incomplete. Expected: 13/32; Found: 975/2304
Bar 96 Stave 2 incomplete. Expected: 13/32; Found: 975/2304
Bar 96 Stave 4 incomplete. Expected: 13/32; Found: 975/2304
Bar 96 Stave 5 incomplete. Expected: 13/32; Found: 65/144
Bar 96 Stave 14 incomplete. Expected: 13/32; Found: 39/2304
Bar 96 Stave 15 incomplete. Expected: 13/32; Found: 13/72
Bar 97 Stave 1 incomplete. Expected: 91/224; Found: 39/2304
Bar 97 Stave 2 incomplete. Expected: 91/224; Found: 65/144
Bar 97 Stave 4 incomplete. Expected: 91/224; Found: 13/576
Bar 97 Stave 5 incomplete. Expected: 91/224; Found: 481/1152

Reply

• Nov 20, 2015 - 15:48

Indeed, the corruptions are almost certainly the cause of the crash, so first order of business would be to understand the source of the corruptions. Right now, as I said, there are a small handful of known issues with local time signatures. The one that seems most likely to be involved here is #51626: Copy/paste to/from passage with local time signatures produces corruption.. Do you remember using copy and paste in any of these passages? Looks like it may well have been used in the clarinent parts in measure 94, for example.

Anyhow, the solution for now will probably be to completely delete measure 94-97, insert new measures, and re-enter the music without the use of copy & paste. There are often other easier ways of fixing corruptions as suggested in another response, but local time signatures introduce enough complications that I would not expect much in this case. Also, in the future, do be on the lookout for corruptions like this and see if you can remember the steps that led to them, and then please report the issue here so we can investigate further.

Ultimately, of course, we need to fix the copy & paste issue. There quite a number of very difficult things about making it work correctly, so more likely we'd end up just disable it, but even that turned out to be non-trivial which is why it hasn't been done already. I may take another look. I have added a link to this thread in the official issue report, so we will have record that someone has encountered this bug.

Reply

• Nov 20, 2015 - 21:31

I did not copy paste those parts on the crazy time signatures. I entered them manually, But, yes, repeating ostinatos I did copy/paste.

Okay, I started deleting all of those crazy measures. But I guess part of the corruption is that the rests don't add up, you can't delete them to get a full rest again, and you can't change the time signature for that measure because the tuplets don't add up, apparently. That and when I go to delete the piccolo and flute parts, the runtime error pops up as soon as I click on it. Have a look:

Attachment Size
North American Suite I, Mvmt I.mscz 77.51 KB

Reply

• Nov 20, 2015 - 22:01

Right, it's not enough to delete the *contents* of the measures. You have to completely delete the measures themselves - Ctrl+Delete - then re-insert new ones. I tried it and it worked to select just the measures that reported as corrupt. When I saved and then reloaded the fie, ni more corruptin was reported. But when in doubt consider deleting an extra measure on either side, or all the way to the start to the end of that section of the piece if necessary.

Reply

• Nov 21, 2015 - 02:24

Wait, nope. I got it ! I just went into the measure properties, changed the "Actual time" to 4/4. And the deleted it. Now is this bug related to local time signatures or is just that every thing got screwed up in my score with the conflicting times one after another ? Because I need these signatures. I can't work without them.

Reply

• Nov 26, 2015 - 16:09

To be clear: local time signatures are defined (in MuseScore) as different time signatures on different staves at the same time. This is currently supported but kind of experimental, and is known to have a number of limitations, plus the bug I already mentioend where copy and paste produces corruption. Right now that is the only such bug we know of. So if you are finding a score with local time signatures being corrupted even though copy and paste is *not* being used, we would need a sample score that is not already corrupt and precise steps to reproduce the corruption. Then we could log an official bug report and investigate further.

Reply

• Nov 27, 2015 - 17:28

It seems that copy-paste shouldn't be used then. That's stupid. You can see all the ostinatos and similar parts I have. It's much more tedious to manually notate every individual note. I suggest you make fixing that bug top priority. I'm done here. You've not helped me at all. I'll just a global time or just rewrite the score manually. Is there a way we can delete this thread for the privacy of my work ?

Reply

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