Unable to save and reopen tuplets correctly

• Oct 23, 2021 - 13:04
Reported version
3.6
Type
Functional
Frequency
Once
Severity
S3 - Major
Reproducibility
Always
Status
needs info
Regression
No
Workaround
No
Project

While one can correctly type in triolets (and mostly any tuplets), reopening the saved file can lead to error messages about incomplete/or overfull bars ("expecting 4/4, getting 36/24" etc.).

Tested on Archlinux, Musescore version 3.6.2

Steps to reproduce:
1) type in Musescore the original score (see image musescore_triolet_original.png), save the file and close it.
musescore_triolet_original.png

2) reopen the previous file, and compare bar 2 (see image musescore_triolet_bug.png). The 4th note has been converted to a silence ("-.").
musescore_triolet_bug.png

This test is from the bass line of the given gp5 file (available for download I do not remember where on the internet, see bar 75, with another similar error at bar 77 for different instrument).

I stumble upon many of these errors when trying to convert gp5 files to musescore : the import is absolutely fine (no error messages, the displayed score is the "same" as the original), but the saved mscz is corrupt. In some cases, the note just disappear or the bar is completely messed up. Sometimes, rewriting the offending notes is OK, but in this case, it is not (unless I did something wrong)

Best regards


Comments

Frequency Many Once
Status active needs info

"type in Musescore the original score (see image musescore_triolet_original.png), save the file and close it".

Do you mean there's an error in opening the GP5 file? It works perfectly for me if I type the score into MuseScore (as apposed to importing it from something). OS: Ubuntu 21.04, Arch.: x86_64, MuseScore version (64-bit): 3.6.2.548021370, revision: 3224f34

To be clear, this shows a sixteenth note followed by an eight note which, together, are played in the space of one eight note followed by 8th note-(tied)-8th note-(tied)-quarter note, 8th rest, 8th note, 8th note. The score is full of them so there must be something odd about those particular bars, possibly in the GP5 file itself.

Attachment Size
!!aa.mscz 5.95 KB

Thank you for the fast reply :
Yes, I can open your file. I think my test file was already corrupt when I tried to correct the bar 75.
and yes, the gp5 file is full of these triolets but oddly enough bars 75 (staff 1) and 77 (staff 4) are the offending ones.

I should have simply described the steps as

1) import the provided gp5 file, and save it as mscz then close it.
2) reopen the mscz and look at bar 75 for bass (staff 1), and bar 77 for lead vocals (staff 4)

I do not know whether this gp5 has something special in it, but I also have many other gp3-4 files with the same problem.

See below screenshots for bar 77 from tuxguitar PDF export in case you do not have it installed (tuxguitar ver 1.5.4)
tuxguitar-bar77.png
and in Musescore (after reloading the imported file)
musescore-bar77-bug.png
There is a silence ("𝄽..") where there should not be (the mscz staff is written one octave lower but the notes pitches are the same)

Attachment Size
tuxguitar-bar77.png 3.39 KB
musescore-bar77-bug.png 8.49 KB