Crash when changing note duration in a staff with a linked TAB staff

• Mar 23, 2021 - 19:42

Severity: S2, crash
Type: Functional
Frequency: no reports found
Reproducibility: Always

Description:
double click on attached file: "nhwy linked TAB issue.mscz"
opens in continuous view
command-f 42 return
using mouse, select last note in m42 (beat 4) in staff 5 (Bass Guitar)
attempt to make a tie with "shift +" keystrokes
immediate crash

Note: Almost any attempt to change note duration in this staff produces this bug.

Expected behavior: notes are linked with a tie
Actual behavior: Crash. Then on reopen, "The previous session quit unexpectedly. Restore session?"

Yes. It opens without corruption in this case. But note this file is probably very messed up. Between my experimentation, not knowing how to set up linked TAB staffs, and frequent past notices of found too many beats in some measures, ...

The problem only happens on the staff with the linked TAB staff.

MuseScore About reports:
OS: macOS 10.16, Arch.: x86_64, MuseScore version (64-bit): 3.6.2.548020600, revision: 3224f34

Mac About reports:
MacOS 11.2.3 Big Sur

nhwy linked TAB issue.mscz
Problem Details and System Configuration .txt


Comments

It crashes simply because the "Bass" part is corrupt (empty measures, from measure 43 etc. - image below)
emtpy.jpg
Remove this part and recreate it to fix this corruption, see: nhwy linked TAB issue 1.mscz

" Between my experimentation, not knowing how to set up linked TAB staffs"
Probably one of the reasons. But I think, unfortunately, there is an issue (introduced recently - as far as I can recall for now) with linked staves and parts in a certain sequence of steps.
I had encounter it a few days ago. This sequence of steps is unusual, or not really supported in first intention (but it did not cause corruption until recently), so I didn't go any further for the moment to understand everything that's going on.

In reply to by nwhy

"That image is how it looks when you open it? Now it is even stranger to me. This is how it looks when I open it."
Your image shows an excerpt of the main score. I was talking about, in my comment, of the Bass part.
Go in this part and look from measure 43 after disabling multimeasure rests (by pressing shortcut "M")

In reply to by cadiz1

Yes, that was a screen shot of the whole score. I can choose to view the Bass part, but I still do not see the same thing you do. There are no multi-measure rests in the Bass part, but I tried disabling them anyway using the m shortcut and I do not see any difference. This is what I see for the Bass part.

Screen Shot 2021-03-23 at 16.24.18.png

Should I download a different version of MuseScore? There used to be another Bass part but I removed it and as far as I know it is no longer in the file. But when I play the score, I can hear something that I can not see. As I mentioned, this file has had a "learning experience" which might be unusual.

Attachment Size
Screen Shot 2021-03-23 at 16.24.18.png 116.9 KB

In reply to by nwhy

There is 2 files in this thread! :)
1. Yours, attached in your first message, ie: nhwy linked TAB issue.mscz (please check)
In this one, there is multimeasure rests - image below - and visible corruption as shown in my first reply when disabling the MM rests.
bass.jpg
2. The same file, in the same reply, but fixed by me after removing and recreating this Bass part, ie: (note the 1, at the end of the "new title"): nhwy linked TAB issue 1.mscz
So, here: score repaired = no more corruption :)

In reply to by cadiz1

Sorry, I mistook the "1". You are right (of course). The Bass part of the corrupt file has either multi-measure rests or empty measures depending on the m option. I see that now, thanks for all your help! I don't understand why the full score view of the corrupt file is different -- it shows those notes if they are not there. I can hear them when I play Bass part in the corrupt file in both settings (with multi-measure rests, the cursor advances as it plays, without them it waits at the end of the good section until the playing gets to the next good section.

When I do "measure properties" on a corrupt measure in the full score view, it looks normal (eg 4/4 not something like 21/17).
Each of those bad ranges of measures I think were caused by me trying to edit the note durations in some way.

I will be able to get back to this tomorrow afternoon. I look forward to looking at this with these new insights.

In reply to by nwhy

After further checking, in fact, I think this is a duplicate of this issue: #313042: Corruption in the parts if linked staff is added afterwards and leading to crash

See also:
https://musescore.org/en/node/313038#comment-1039809

With another scenario a bit more developed - which would be more like the present case of this thread - so, with the intent to display TAB et standard staff both in parts, I get the same massive corruption (empty measures)
Is it really useful to give the steps? I can do it, but apparently the roots of the problem are the same: when you change the staff type (eg Standard for TAB), this change is not updated in an already created part.
And if, on top of that, you do the same operation again to display two different staves (TAB and standard) in the parts, you multiply the problems and open the way to all possible corruptions!

EDIT: Having said that, as far as I remember, this is a fairly recent regression. This kind of operation was known as a workaround - this was not without ("theoretical") risks, but it worked anyway - for displaying standard staves and TABs in parts, as Guitar Pro can do (and because numerous users - frequent request - would like obtain the same thing easily, "natively")

This is no longer possible, something has been broken (I have an idea, but I'd have to check first)

In reply to by cadiz1

Yes. This was my first attempt at a TAB staff in MuseScore. I know I was following the instructions from
https://musescore.org/en/handbook/3/tablature#staff-tab-existing-score
I was definitely adding new staff(s) to an existing score. At this point I do not remember if I tried to add a linked TAB staff to an existing Bass clef, I might have tried that and then added a new linked pair instead.

Copy, paste, add and delete worked, which made me think the crashes were from duration changes. At that point I did not know the file was corrupted. There were many times I got the "measure incomplete. Expected 4/4; Found 11/8" messages. Which I cleared good enough to load, but perhaps did not clear totally.

Testing your fixed version of the file, it is clear that the corruption was the cause of my crashes, because note duration changes now work great.

Thanks again.

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