Should accidentals move with notes/chords?
See #25648: Accidental does not move with manual adjustment to note/chord for background.
I am trying to decide what the best thing to do is if a note has an accidental and then the horizontal position of that note is moved manually - either by moving the note directly via Edit mode, or by moving it or the chord containing the note via the Inspector.
In the case of a single-note chord with nothing in any other voice in the staff, it feels natural to me that the accidental should move with the note. But as soon as we start talking about chords with multiple notes or chords in multiple voices, things get a lot fuzzier. Accidental positioning is complicated in that it is not specifically tied to the position of any given note or chord - they are computed relative to all notes of all chord in all voices at that time position. Accidentals are all stacked to the left of an "origin" point that is the leftmost note in any chord in any voice on that staff. Any manual positioning that changes this origin point would basically move the whole stack; any manual positioning that does not change the origin point would have no effect.
My hope is that with the layout improvements in 2.0, there will be virtually no need to move notes or chords just to fix layout issues. Users will only move notes or chords for unusual special situations, and there is no way to predict whether having the accidentals move will feel right or not. So maybe leaving it alone - where moving note/chords has no effect on accidentals - will be OK. But I'm opening it for discussion.
Comments
(Sorry, I missed this post; apologies for the late reply.)
With the recent improvements, needs for manual adjustments are much reduced indeed.
Just for completeness, one case comes to my mind which is (not exceedingly but) decently common and still requires manual adjustment: when stem-up and stem-down notes / chords alternate in sequence.
The best practice in this case is to space notes not with regularly spaced note heads (which is the out-of-the-box MuseScore layout and indeed the 'normal' practice), but halfway between regularly space note heads and regularly spaced stems, more or less like this:
NOT FOUND: 1
The example uses single note chords and does not has accidentals, however it is easy to extend. In this case, I think I would expect the accidentals of each chord to 'follow' the displacement of their chord, otherwise accidentals of up-stem chords (which are moved to left) would clash with it and accidentals for down-stem cords (which are moved to right) would become detached from it.
Just my 2 cents...
M.
In reply to Just one special case... by Miwarre
OK, that's one case where moving the accidental stack would make sense. The case that brought this to my attention is another: in music that requires more than 2 independent voices on a single staff, we often need to offset chords in voices 1 & 3 (or 2 & 4) so they can be seen distinctly. Most often, this would probably be done by moving one of the chords to the right, leaving the other chord in its default position, and we don't need accidentals to move in that case. But in cases where collisions are resolved by moving one of the chords to the *left*, you would want the accidentals to move as well. My proposed change deals with this - moving one of the chords to the right would not affect the accidental positioning because the left-most note is the only determinant. But moving one of the chords right would move all the accidentals to make room, as would be desired.
The proposed change is easy enough to implement; see https://github.com/musescore/MuseScore/blob/master/libmscore/layout.cpp…. Feel free to try it out, and if you think it's the way to go, I can submit a PR (or you can :-)
In reply to OK, that's one case where by Marc Sabatella
Also if voices 'overlap' (voice 2 or 4 being higher than voice 1 or 3), I think. Not sure how 2.0 handles this, but in 1.x is was neccessary to nudge notes to be able to tell which stem belongs to what note head
In reply to Also if voices 'overlap' by Jojo-Schmitz
Nudging shouldn't be needed any more for overlapping chords if only two voices are involved. See http://musescore.org/sites/musescore.org/files/issues/seconds-4.png (and ignore the possible slight glitches in how the stems attach to noteheads; those were zoom artifacts).
In reply to Nudging shouldn't be needed by Marc Sabatella
OK, looks good.