Ties are poorly laid out inside clusters

• Jan 4, 2019 - 16:13

See some examples attached

Attachment Size
MuseScore3_2019-01-04_17-13-38.png 23.77 KB


In reply to by Marc Sabatella

I'm not sure if it's really hard to detect. Don't focus on the lines themselves, cluster notes are a good heuristic. Typically if it contains 2nds (alternating left and right) it usually produce bad ties. OK maybe the rule should be a little more nuanced but it doesn't seem impossible.

In any case, even just a checkbox to get "basic" ties would be golden.
I'm spending crazy amounts of time fixing my ties manually, it's really time-consuming.

In reply to by jods

"I'm spending crazy amounts of time fixing my ties manually, it's really time-consuming."
I agree: currently I am reviewing songs with piano accompaniment, and fixing the layout of clustered ties in the piano part is ... well, repetitive and tedious. And anything repetitive should in theory be open to a programmatic solution? I would have a go, but sadly I never learnt C++.
The situation is even worse where dotted clustered notes are involved, because the left-hand anchor can collide with the dot in too many cases.

In reply to by jods

Would you agree with this assessment:

If the chord is nothing but two notes a second apart, and there are not multiple voices, the defaults generally work pretty well. The ties are in opposite directions and stay out of the way OK. Things get iffier if there more than just the one second, or there are multiple voices (which force stems in the same direction. At that point, it really gets more difficult to predict.

So the best rule I could propose might be, use the current rule for chords with no seconds, or for chords of only two notes with only a single voice. But for chords containing seconds and having either more than two notes or multiple voices being present, use the alternate rule where ties attach outside of the full chord.

Do you still have an unanswered question? Please log in first to post your question.