No way to create full measure rest in voice > 1

• Aug 17, 2014 - 17:24
Type
Functional
Severity
S4 - Minor
Status
closed
Project

1. Open attached score (produced in 1.3).
2. Click on voice 2 note.
3. 'Delete'.

Result: The voice 2 bar rest is off-centre.

Multi-voice bar rest off-centre after deleting note or producing rest amounting to bar value.png

Note: It is also visible in a 1.3 score if such steps are performed in that version.

Using MuseScore 2.0 Nightly Build (01b6a81) - Mac 10.7.5.


Comments

This is a bit problematic, and we need to decide how to best handle this.

The reason it is offcenter is that it is a whole rest (a regular four-beat rest) as opposed to to a "full measure rest" (identical looking symbol but can be used to indicate an empty measure of any time signature). Same thing you see if you enter a whole note into voice 1 then delete it. It's totally proper that whole rests be aligned left. Only full measure rests are supposed to be centered. And for the measure in question, it's proper to use a whole rest as is the case here - a full measure rest would be inappropriate and harder to read.

Now, for voice 1, you can create a full measure rest by selecting the whole rest pressing "delete" - the whole rest is removed, leaving the measure empty, so the full measure rest automatically takes its place. but for voice 2, deleting a whole rest just deletes it - a full measure rest is not inserted in its place. Nor would you normally want it - it's got to be possible to completely removed the contents of voice 2. Whereas with voice 1, it's never allowed to be completely empty. Any otherwise empty measure is automatically filled with a full measure rest in voice 1.

There is no way I know of currently to create a full measure rest in voice 2 except as a side effect of a bug - #28886: Full measure rest created in next measure if changing length of chordrest in voice 2,3,4 creates rest at end of measure. And the layout code doesn't currently deal with full measure rests anywhere but voice 1.

The question to me is, *should* this be supported? There *are* legitimate use cases where you would want a full measure rest in one voice but have notes in another voice - this example just doesn't happen to be one of them. But a part with note notes except cue note might be a legitimate use case. One workaround would be to make sure the cue notes are in a a voice other than 1, so voice 1 can hav the full measure rest.

I'm trying to think if there are any other good use cases that suggest you *should* support full measure rests in voices other than 1. If so, we need a way of creating them, and we need to make the layout code work with them.

Title Multi-voice bar rest off-centre after deleting note or producing rest amounting to bar value No way to create full measure rest in voice > 1 (and wouldn't lay out correctly even if you did)

The usual convention is, don't do this if the different voices on the staff are to be read by one musician. So you don't do it for drum set. But it can be done if the voices represent independent parts for separate musicians. This might happen if you use one staff for the music for several percussionists as opposed to a single drumset player. Or, in orchestral music, two flute parts on one staff. Even so, it's just as common to simply omit the rests or use regular duration rests (eg, whole rest in 4/4) which would be left aligned. but at least some sources do recommend use of measure rests here, so we *should* support it.

Workaround for now is to place the empty part in voice 1, but this might require manual flipping of stems or adjustment of vertical position of rests.

We currently have a command to convert an empty measure into a full measure rest if it is not already. This could be enhanced to work on voices > 1. And then we'd need to fix the layout.

Underquark suggests a simple but effective method: exchange voices 1-2, then back. And with the layout working as expected, I'm going to consider this closed. I suppose one could still submit a feature request for a more easily discoverable method, but the underquark method at least doesn't seem like a bug.