Wrong playback of trills with ottava

• Mar 27, 2019 - 11:47
Reported version
P1 - High
S3 - Major
needs info

Windows 7, MuseScore 3.0.5

1) On an empty 4/4 measure write a quarter note and select it
2) Apply a trill from the Lines palette
3) The trill covers the whole measure duration instead of the intended note.
4) The duration of the played trill is correct
5) Add a second note without shortening the trill
6) The second note plays as a trill as well
7) Remove both notes or delete the measure content
8) The trill is still there, hovering over the silence
9) Seemingly, for scores originally created in 2.3.2, the trill does not move with the note, so if the note is moved upwards to ledger lines, eventually the note conflicts with the trill (See attached file). I cannot reproduce this problem in a score created in MuseScore 3.0.5. If the cause is confirmed, it is a case of forward compatibility. This would be a very important bug, since new versions should be able to open correctly and convert to the new format. Otherwise one cannot be sure a score will open cleanly with future versions.
10) Whenever there is an ottava present, even if is a transposing instrument as the piccolo, trill playback alternates notes from two different octaves. UPDATE: I made a mistake: I was testing both 2.3.2 and 3.0.5 and got confused. Ottava Trills work as expected in 3.0.5, even with a score originally created in 2.3.2. The reported behavior regarding ottavas corresponds to 2.3.2, not to 3.0.5. My apologies.

Attachment Size
test_trill_1.mscz 10.57 KB


Title Awkward behavior of trills Wrong playback of trills with ottava
Status active needs info
Priority P1 - High

This seems like a bunch of unrelated things, so I'd encourage to submit separate issues. How are you adding the trill? If you correctly select the range first then double-click, it gets appleid to just that range, but otherwise, lines get added to end of measure by default if you don't select an endpoint. This is by design, it's what you might expect for some kinds of lines some times, but in general, indeed, we won't be able to read minds and know the expected result all the time. We could conceivably special case trills, indeed. That's worthy of a "Suggestion"

Lines remaining after deleting notes is also by design, it's what you want often but indeed not always. Again, we could conceivably special-case trills, so that could be a separate "Suggestion".

Trills didn't respond to note position in 2.3.2 - overlaps could and did happen - so actually, what you descirbe as the current behavior could be seen as if the goal is to mimic layout of 2.3.2 scores. But, this isn't what is actually going on - it's that we don't do this form of automatic placement in continuous view - again, by design, as with not allocating additional space for lyrics.

The octave playback seems like a true bug, so I've re-titled this issue accordingly, but I can't reproduce, so we'd need a sample score for that.

First and most important: 10) was mistaken, I have updated it. See above. Sorry that the incorrectly reported behavior caused the original title change to be merged with a prior report that was valid for 2.3.2 but not for 3.0.5
The trill has been added either by dragging the trill from the palette or selecting the notehead and double clicking on the pallete icon. In both cases the trill lasts through the entire measure, but if I select the range instead of the note, it lasts until the note ends.
Most trills apply to single notes, so it is a question of statistics. It is a similar case as the one-note slur, which as expected extends only to the next note. Ottavas have the same behavior, but in this case usually ottava is aplied to a note stream.
As to non automatic positioning of trills in continuous view, it seems inconsistent with other objects. For example, if there is an instrument change and one moves the note upwards, if too close, the instrument change text starts to move. Why trills and, by the way, also ottavas behave different during continuous view?