[trunk] With some time signatures and irregular measures, properties of foll. measures are corrupted after saving and re-loading
Context: trunk rev. 5503 (but occurring since a while), compiled under Win 7 with latest Qt SDK and Qt lib 4.8.0
Steps:
1) Create a new score with 3/8 as time signature and pickup measure
2) Enter a few notes
3) Check properties of any measure: "nominal" duration is correct: 3/8 (see attached before screen shot)
4) Save, close and re-load
5) Check properties of any measure: "nominal" duration is incorrect: 4/4 (see attached after screen shot)
Notes:
For this to happen, 3 conditions are required:
1) time sig different from 4/4 (or from mathematically equivalent to 4/4, like 2/2)
2) a 'short' (or pickup) measure
3) some notes after the 'short' measure
This happens either:
*) if pickup measure is initially created in the "New Score" wizard
OR
*) if pickup measure is created afterward, by reducing the actual duration of a measure
and both:
*) if the 'short' measure is at the beginning of the score
AND
*) if it is somewhere else (for instance at the beginning of a seconf movement)
Marked as critical as this leads to data corruption. Incorrect nominal duration aftects beaming and new measures appended or inserted. There also seems to be no work-around nor a way to go back to normal.
Attachment | Size |
---|---|
trunk_pickup_corrupts measures_before.png | 41.94 KB |
trunk_pickup_corrupts measures_after.png | 42.85 KB |
Comments
By 're-load', do you mean re-open?
I do.
M.
P.S.: better to keep [trunk] indications in titles for trunk bugs: branch has not been discontinued yet and any error reported by users of the release version will refer to it; so I think it is better to keep the distinction.
Okay - just because of this: #11620: [Trunk] Reload causes crash
I think lasconic instructed me to remove the trunk title from issues, because there will be no more versions on the branch.
I can not reproduce this. Can you do a retest and close it if it is no longer relevant?
Have just started using this after many years of using PrintMusic and am delighted with many things for piano that were hard or impossible in that programme (beaming, tremolos etc).
However, after several bars of 4/4, I created a 7/8 bar, followed by a 2/4 bar (which I hope is a temporary measure...don't you wish students would learn to count to 4...?!), then back to original 4/4.
My problem is that after the time change bars, the '4/4' bars add up to 9 quavers, then 8 (good), then 1 quaver. After those 3 bars, it seems to correct itself.
Thanks,
Antonia
I hope this message is in the right place...don't know if reloading/saving made any difference.
To add further info: the unintended 1/8 bar is letting me put 8 quavers into the left hand. It is letting me add a 4/4 time signature to every bar after the time changes. (Maybe I should stop making it worse...!)
Sounds like you managed to corrupt your score. Please attach it here, so someone can have a look.
Or just delete (Ctrl-Del) the affected measures, insert new ones and enter your music again.
I did that, and didn't dare reinsert the 7/8, but it's happy again.
Thanks!
Nobody came up with ways to reproduce it in the current code, so let's put it to a rest
Automatically closed -- issue fixed for 2 weeks with no activity.