Key Signature appears too many times in Parts in 3.5 Beta
In Musescore 3.5, in a Part, the key signature appears in multiple places on the staff. It should appear only at the beginning of the staff. In the Score, it appears correctly at the beginning of the staff only. See screenshots. Musescore file is attached.
In Musescore 3.4, both the Part and the Score key signatures behave correctly, using the same Musescore file.
I'm using Musescore 3.5:
OS: macOS 10.15, Arch.: x86_64, MuseScore version (64-bit): 3.5.0.27580, revision: b5add95
Attachment | Size |
---|---|
Test Key Signature PART.jpg | 226.26 KB |
Test Key Signature - SCORE.jpg | 204.87 KB |
Test Key Signature.mscz | 31.66 KB |
Comments
I don't see this in 3.5 Beta on Windows 7
In reply to I don't see this in 3.5 Beta… by Jojo-Schmitz
Thanks for checking it. I'm using a Mac running Catalina 10.15.5.
Janet
Was this score originally created in an older version, maybe 3.3 or 3.2 or earlier? There were definitely bugs fixed a while ago where sometimes these phantom key signatures could get inserted. The bugs were fixed, but any key signatures added this way would still remain, waiting for a chance to reappear, even if the score is opened in a more recent version then saved.
Anyhow, I can't reproduce any problem using 3.5 beta. I opened the score then click the part tab, and it looks as expected for me. Is there something else I should need to do to see a problem?
In reply to Was this score originally… by Marc Sabatella
Hi, Marc
Hmm. The test score was created with 3.5 beta (installed July 4) specially for this test. And this morning, the problem behavior isn't there any more in the test file.
But then, yesterday (July 15), AFTER I reported the problem, I updated to the current 3.5 Mac download ... so maybe something in the new download fixed the problem.
Do you have any workarounds for the problem in my existing scores? Adding parts for instruments that have never had a part before work fine. Parts for instruments that had the problem still have the problem.
My existing scores were created with 3.4.2 (downloaded February 2020) or 3.5 (downloaded July 4). I have both versions installed. My 3.4.2 is an update to whatever I installed over a year ago.
Thank you for your quick response.
Janet
PS Next time, I'll update 3.5beta before reporting anything!
In reply to Hi, Marc Hmm. The test… by Janet E.
As far as I know, there is no update, the Beta that was available on July 4 is the same as what is there now. But what is possible is that whatever the problem is, the mere act of saving and reloading the score is what fixed it.
The bugs mentioned were fixed long before 3.4.2. Scores created in older versions that have that particular problem can be corrected by turning off courtesy key signatures, then saving, then reloading, then turning them back on. But whatever you are seeing here seems to be some new problem if the score was created from scratch in 3.5 Beta, so we'd really like to understand more about how to reproduce the problem.
In reply to Hi, Marc Hmm. The test… by Janet E.
Hmm, indeed, confirmed, there is a problem (3.5 beta and current nightly, well July 10 for now, I will update soon) at least with piano instrument (2 staves?), in parts, and related to system breaks at first glance.
From scratch
1) Load this new score, piano template, in D (2#) : Piano test.mscz
(and system breaks every 4 measures)
2) Generate part (all parts)
3) Toggle in part and disable MM rests
Result: unexpected key signatures (apparently at each system break, apparently because seems a bit different eg with every 6 measures eg)
Further investigation needed. I check ASAP where the change occurs (between 3.4.2 et 3.5 beta)
In reply to Hmm, indeed, confirmed,… by cadiz1
Aha, yes, I see, thank you for the steps to reproduce! Knowing when the problem was introduced would certainly be useful as well, so if you're able to get that info too, great!
In reply to Hmm, indeed, confirmed,… by cadiz1
Same result from scratch with last nightly, July 14: 0c1e286
EDIT: curious, I receive this result from scratch and mentioned steps:
BUT, after save/reload, the issue is gone!
See the test score at this step: test 2 piano.mscz
and image:
In reply to Same result from scratch… by cadiz1
In April.
I continue to narrow down, and then report it in the Issue Tracker.
Edit: between 20 and 25 April...
In reply to In April. I continue to… by cadiz1
April 21 the most likely.
I try to be more precise yet soon.
In reply to April 21. I try to be more… by cadiz1
Confirmed for April 21.
Somewhere between this nightly (works, as expected): 333752a
and this one (unexpected) : eeba202
In reply to Confirmed for April 21… by cadiz1
I only see two commits (related to Piano Roll and about delete text with Ctrl + BackSpace) between the two previously mentioned nigthlies.
So I suppose an unexpected effect of the last mentioned (Voice - Parts ) ?
In reply to I only see two commits … by cadiz1
Yes, 3 potential candidates, 2 of which look pretty unrelated
In reply to I only see two commits … by cadiz1
Last precisions before report.
Seems to not occur with a single staff (eg Flute).
With organ (3 staves), I see only the problem in the top staff.
And so Grand staff/piano, on both staves.
EDIT: And as said previously, the problem solves itself after Save/Reload.
In reply to Last precision before report… by cadiz1
Done: #307911: Extra key signatures appear unexpectedly in parts from multi-staff instruments, and fix itself on save/reload
In reply to Done: #307911: Key… by cadiz1
If confirmed for this candidate, eeba202, it would be as a follow-up of these two fixes: #291688: Time signature change doesn't appear in parts which don't have voice 1 and/or #285855: Creating part from voice 2 does not respect key changes
In reply to Hmm, indeed, confirmed,… by cadiz1
Hi, Cadiz, thank you for taking this issue and running with it! If there's anything I can help with, please let me know. I'm on Mac Catalina 10.15.5.
Regards,
Janet
In reply to Hi, Cadiz, thank you for… by Janet E.
We have identified the root cause of the problem and have a pending fix, hopefully it will be in the final release. Meanwhile, the good news is this problem fixes itself on save/reload.