Ties not drawn properly with line breaks
Before opening a bug report, I would like to have your opinion about the following two situations.
Open the attached score (created from scratch with MuseScore a92d346 on Xubuntu 14.10).
Problem 1
At the end of the first system, there's a chord with ties:
Ties extend to the first measure of the second system. See how ties are weirdly drawn:
It seems that the second is the culprit.
Problem 2
At the end of the second system, there's another chord with ties:
Ties extend to the first measure of the third system. See how the middle tie touches the bass clef:
| Attachment | Size |
|---|---|
| chord-1.png | 2.8 KB |
| chord-2.png | 6.84 KB |
| chord-3.png | 2.18 KB |
| chord-4.png | 6.58 KB |
| ties.mscz | 1.8 KB |
Comments
Ties in chords with seconds (your first case) are not great with or without a line break. But indeed, it looks worse when there is a line break. And the whole notes in your example make it worse - the noteheads are larger and offset from normal notes, and the ties aren't accounting for that correctly. That's a bug worth reporting. I should be able to fix that.
The second example is maybe more of a judgement call, but also worth filing. Do it as two separate bugs, since the fixes might be separate.
In reply to Ties in chords with seconds by Marc Sabatella
Done:
Case 1: #37646: Ties in whole note chords with seconds not properly drawn across system break
Case 2: #37651: Ties in chords with line breaks touch the clef
Also, I wonder if there could be some improvement with the algorithm deciding if a tie is drawn up or down. See this example:
I don't know if this kind of automatic detection is on the scope of MuseScore or if it's expected that users will manually make adjustments in such situations:
In reply to Also, I wonder if there could by jpfle
This is a very tricky area. There are a number of competing algorithms advocated by different authorities for deciding tie direction on chords. None produce optimal results in all cases. So while I'm sure there is room for improvement, it is indeed to be expected that in some cases you may need to override the default.
FWIW, the usual rule that is followed here is, ties goes in opposite directions when there are seconds. It's a good rule most of the time, but produces bad results here because it conflicts with the tie in the next note. Other algorithms produce nicer results here but fail in other cases where ours does better.