Significant Play Delay with Polyphonic Tuplets 2.X->3.0
Musescore 3 (nightly build) chokes on simultaneous multiple tuplets when starting playback. In Muscscore 2, the attached polyrhythmic two-bar score starts within 1 second after hitting play on my system, whereas, in Musescore 3, it takes up to 5.
To reproduce the issue, load attached score, start play; compare start times in both Musesore 2.3 and 3. (I am assuming that the play delay is due to how Musescore 3 parses polyphonic tuplets as the score is only two bars long.)
My system:
OS: Windows , Arch.: x86_64,
MuseScore version (32-bit): 2.3.2, revision: 4592407
MuseScore version (64-bit): 3.0.0, revision: d3d0f87
I reported the bug, initially confirmed by Marc Sabatella, twice in the forums:
"Musescore 3 (alpha2) starts slower than Musescore 2 with polyrhythmic tuplets"
https://musescore.org/en/node/278063
Resubmitting bug report: "Musescore 3 impractical to use due to start play delay."
https://musescore.org/en/node/278289
| Attachment | Size |
|---|---|
| Groove_Matrix_Break_Down__offbeat_Nov2018_0.mscz | 11.81 KB |
Comments
The score is over 11 hours long. No doubt MuseScore will take a couple seconds to create internal structures to represent it before playback.
The score itself is very short, just 2 measures, 10 staves and 60 BPM (not 120 as the half note = 60 implies)
It plays parctially endless only due to the 9999 play count
Still it plays normally in 2.x
reducing the play count (to e.g. 9) doesn't make playback starting faster in master
In reply to The score is over 11 hours… by [DELETED] 5
Hi lasconic, I am confused about the title change of this bug report to "...with extremely long scores.". As Jojo mentions, the score is extremely short, only 2 bars--4 seconds long. Regardless, Musescore 2 handling of this score is fivefold faster than Musescore 3 on my system.
reverted title, as the length (via play count) has got nothing to do with this
The issue appears not to be the number of simultaneous tuplets, but with how Musescore 3 converted the original Musescore 2 score. Perhaps there are coding errors in the original score. A similar score, with the same number concurrent tuplets, created in Musescore 3 does not have the delay (see new attached file).
(In the attached file "Test Tuplets musescore 3", I wasn't able to make the name of each staff identical to the original because the selectable area is so narrow on some staff lines in Musescore 3 that the staff properties right-click option isn't accessible-has this right-click bug been reported? Also, the instrumentation is not the same, but the arrangement of tuplets is.)