Change the time signature remove the score elements

• Mar 17, 2015 - 17:53
S1 - Blocker

MuseScore 2.0 RC
Change the time signature removed the system jumps, page breaks sections breaks... and others.

Either this partition with sections breaks :

Disparition 1.png

Add time sig. between second and third :

Disparition 2.png

Add time sig. between first and second :

Disparition 3.png

If change time sig. of the score : system jumps ( page breaks ) and section breaks are removed. Note : double barline also are removed ( and corrupts bars ? )

Disparition 4.png

Sections fuyantes.mscz

Attachment Size
Disparition 1.png 15.67 KB
Disparition 2.png 15.58 KB
Disparition 3.png 14.93 KB
Disparition 4.png 14.93 KB
Sections fuyantes.mscz 4.06 KB


This is a known and to some extend unavoid limitation of time signature changes. Now that measures are re-written to reflow notes across the barlines, usually changing the number of measures as well as which notes are in which measure, there is no particular way to know where it would make sense to add the breaks - the old measures are simply gone.

However, it would at least make sense to stop rewriting time signatures at a section break. See #47356: Change of time signature removes subsequent section break.

Initially it is not about to change a time signature or change the number of measures, but having to repair a file having undergone corruption when loading on MuseScore 2.0RC ( discordance between the time signature displayed and real time signature ).

And that is perhaps more serious than you think because it seems to be related to this problem:

If the following sections have the same time signature is a more distant section which can be corrupted without being noticed with big inconvenience to the reopening of the file.

View thread in the French forum :

Status (old) duplicate needs info

The problem in #51146: The change of a time signature before an inserted frame causes corruption is not related and in any case is already fixed.

You speak of corruption, but you haven't posted any steps to reproduce any sort of corruption, nor does the image you posted appear to show any. If you have some evidence of corruption, please post the precise steps to reproduce ASAP, because time is running out to fix critical bugs for 2.0.

But again, let us be clear: the fact that changing time signature means barline are no longer in the same place, and that means breaks and barline lines can no longer make sense. We could try to *guess* which of the new barlines - barliens that did not exist before - should perhaps get breaks or double bars to make up for the ones that no longer exist - but there is pretty much no way we'd guess right, so it would be twice as much work to fix. You'd have to delete the wrong guesses we inseerted before inserting the correct breaks and double bars. I don't think it would save you time on average. So, as a feature request, sure, we could implement as "guess where the suer wants to insert new breaks and double bars" facility, but lack of this feature is not a bug, and I suspect if you had the feature you'd end up cursing it as often as you thanked it.

The only real bug here is that time signature change does not stop at the section breaks. And, if you have some sort of corruption that you can post precise steps to reproduce.

I think we do not understand (bad translation with Google no doubt?). I just want to point out that a correction must not lead to other corruptions. A section must be autonomous for the indication of time (we agree on this point) and number of measures to allow precisely a time signature change (and therefore the number of measures) possible without compromising other sections.

More simply: with MuseScore 2.0RC in the presence of section breaks, it does not seem easy to restore a corrupted partition.

I've looked at this a little. Stopping at the section break is the easy part. Slightly tricker is reinserting the break on the new last measure of the section, and slightly trickier still is re-establishing the time signature after the break if there wasn't one already. Probably not too hard, though, if you know what you are doing.

Note that this issue occurs even if the time signature change is from (e.g.) 4/4 to 2/2, where there is no note re-flow and no ambiguity about where the elements should occur after the change. Given that changes like these are probably the most likely time signature changes one would make in an existing score, perhaps an effort should be made to preserve the score elements in such cases.