Playback cursor does not follow complex tuplets
Reported version
3.x-dev
Priority
P1 - High
Type
Graphical (UI)
Frequency
Many
Severity
S3 - Major
Reproducibility
Always
Status
closed
Regression
No
Workaround
No
Project
As attached V3.1 score, when playing bar 17, playing bar and playing note focus dispersed from beat 3. This can be visually traced by slowing down the play speed to 10%, as screen captured.
And it looks the playing bar is falling behind the current playing notes.
The same score works well with MS2 for this same bar.
Attachment | Size |
---|---|
Playing focus dispersed.JPG | 175.75 KB |
Nocturnes, Op.9 (Chopin, Frédéric) No.2 - Violin & Piano V3.1.mscz | 43.83 KB |
Nocturnes, Op.9 (Chopin, Frédéric) No.2 - Violin & Piano V2.mscz | 47.3 KB |
Fix version
3.1.0
Comments
I agree, the playback cursor lags when the piano in invisible. Also, the two instruments play at different rates as shown by the colored notes and the playback cursor. I don't see any immediate reason for this happening. Perhaps the 30-let is causing it.
When you make the piano visible the notes line up and play properly.
In reply to I agree, the playback cursor… by mike320
You are right, do you have an quick remedy to fix it before updates with piano invisible?
I might have to go back to version 2, it works well there.
In reply to I agree, the playback cursor… by mike320
Bar 3, 15, 16.... have the same problem, it's XX-let which causes it or hidden piano. It looks there is no quick remedy before bugs get fixed.
The only fix I know of is to make the piano visible for playback.
In reply to The only fix I know of is to… by mike320
We wait then for updates.
Actually the invisible piano part isn't the issue, same thing happens if you simply delete it. Problem seems to be in the calculation of where the playback cursor actually is. Presumably something was that missed in the tick to fraction conversion a couple of months ago.
Meanwhile, unless I am missing something this is an onscreen visual glitch during playback only, so I've adjusted severity and set priority accordingly.
You seem to be missing something. When the piano is invisible, the notes are played at the wrong time. You can hear the beat 8 piano notes playing while the cursor is still on the 30-let and playing notes. I hear them start about the same time as the G-natural with a 3 above it on beat 5.2. These notes should not start until after the dotted 1/4 note starts on beat 8.
OK, sorry, I had my sound off (working at the library), and the description seemed to be suggesting the issue was with the cursor, not the actual playback itself.
In reply to The only fix I know of is to… by mike320
The other fix I found and did: rolling back to 3.0, and it works perfectly well.
Right, it's a regression in 3.1 beta almost certainly caused by the fraction change I mentioned. I now suspect it's actually caused by the same underlying issue as #287883: Sextuplets display as triplet, corruption on copy/paste and as such is already at least partially fixed in master. I'm not 100% sure, because I can reproduce the problem with the score in master - the cursor issue only, playback is actually fine - but I can't reproduce it if I copy/paste the tuplet to a new score. If the issue were only visible using the 3.1 version of the score and not the 2.3.2 version, I'd consider the matter closed, as I don't think it's reproducible from scratch. But the fact that there is a cursor issue working from 2.3.2 is still a concern and needing further investigation. There is obviously something going on, but I'm having trouble pinning it down.
I do not believe that this has anything to do with fractions. It is not an issue in 3.0.5 because 3.0.5 does not include the fixes for #281024: Invisible staves inappropriately factored into note spacing and #286068: Playback cursor not tracking notes on invisible staves.
It is true that if you delete the piano part, this issue persists. But if you then save and reload the score (without the piano part), the problem goes away.
Hmm, maybe you're right -as I recall the playback cursor position is sometimes based on segments, sometimes calculated algorithmically. BTW, the second issue you mention wasn't an issue in 3.0.5, it was an issue caused by my fix for the first. But anyhow, still I don't why the issue exists in master, then. I guess what this says is my fix for #286068: Playback cursor not tracking notes on invisible staves did not do the full job.
See https://github.com/musescore/MuseScore/pull/4984.
Fixed in branch master, commit d3db00d217
fix #288548: Playback cursor does not follow complex tuplets
Fixed in branch master, commit 46480efd8b
_Merge pull request #4984 from mattmcclinch/288548-playback-cursor
fix #288548: Playback cursor does not follow complex tuplets_
Automatically closed -- issue fixed for 2 weeks with no activity.