Change layer of note?

• Mar 8, 2011 - 03:07

I don't just create new scores, I also edit some score in musicXML format. I can't seem to figure out how to change the layer of a note though, which is very necessary for editing. Is this possible with musecore 1.0?

Thanks, Cory


In reply to by darkNiGHTS

I guess the question would be, what you think changing the voice of just one note would accomplish? Even if were possible, it would disrupt the existing contents of both voices, probably producing results other than whatever it is you had in mind that made you want to try it. I'm struggling to imagine a scenario where such a feature would be useful. If the goal is simply to flip the stem of a single note, you can do that directly (the "X" key) without changing voices. I guess maybe if you have independent single-note lines in two voices and except for one two-note chord in one of the lines, and you've decided it would be better to move that note to the other line to have the two note chord there instead? I can't imagine that would come up very often, and it's simple enough to just delete the note form the first chord and add it to the second as it is.

In reply to by Marc Sabatella

It comes up a lot more often than you might think. In classical guitar music, you very often end up using three voices. Especially in Bach. Also, I'm sure I'm not the only one who imports music in the musicXML from other formats and the voicing isn't always correct in these sources. The score that I just edited one page of, I just had to delete 10 measures and redo them because I couldn't change the voice grouping.

It would be nice to be able to right click on the notes and have a menu that says "voice" and you can choose which layer.

In reply to by darkNiGHTS

I'd be interested to see the example. I could indeed imagine it coming up more often after MusicXML import if you have different ideas than the person who created the file in the first place. But I'd still be suspicious that it wouldn't be nearly as simple as you might hope. As I said, moving one note from voice to another would have a ripple effect through the entire measure in both voices that could be implemented any of several ways (eg, the note removed from one voice might be replaced with a rest or by extending the previous note or by shifting the subsequent notes to the left). I find it kind of unlikely that a program would guess the particular results you hoped for often enough for such a feature to be useful. Could you show some examples of what a measure looks like now, and what you would expect it to look like after changing the voice of just a note or two? Maybe the 10 measures you referred to in the guitar piece?

I ask about this because, although I have yet to do any development work on MuseScore myself, I *am* playing around with some ideas for plug-ins that have to do with voice manipulation, and I am trying to get a sense of what would be useful and also what would be possible.

In reply to by Marc Sabatella

It's not as much "guesswork" as you might suspect, it's very simple. All the program needs to do would be add rests in place of the voice that was just moved so the layout wouldn't be changed. And if there are notes in the layer that you are changing it to, it wouldn't let you of course because that would either involve guesswork, or a very complicated options box. But simply moving a few notes from the second voice into the empty third or fourth voice should not be too difficult.

I can't show you the piece I was talking about because it is a composition of a friend of mine, but I can show you an example with another piece. I'm having one problem creating an example though, I make the top line, and then I click "Voice 2" to add notes to voice 2, but it is still adding notes to voice 1. This is a completely blank document too.

EDIT: I closed it and reopened it and now it's working correctly. I'm making the example now.

In reply to by darkNiGHTS

You do know that if you don't change to voice two *before* you enter/copy in the notes that it won't enter the notes in as a second voice-it will enter them as voice one.

However, you could also simply select the measures you would like to change and go to or whichever voice you would like to switch from/to.

In reply to by rj45

Yes, I am aware of that, no matter what I pressed when, it would only put notes in voice 1. I quit the program and reopened it and now it works fine.

In the file examplevoices.png, there are three lines from a Bach Violin Fugue that I am playing on the guitar. The first measure is just there for reference. Let's say I made the mistake of placing the C in the top voice along with the rest of the top line. I would right click on the note, go to voice-> voice 2 and it would create a rest in voice 1 where the note used to be and place a note of the same pitch and value in voice 2 and add rests afterwards. I would change the value of the note to a quarter note and be done.

I can think of a few more examples if you'd like.

Attachment Size
examplevoices.png 26.33 KB

In reply to by darkNiGHTS

I still don't really get it. If the problem is just correcting occasional mistakes, there is also undo for that, and as I said, it's simple enough to make those changes by deleting the note from the old layer and adding it to the new. Second, while in *this particular case* replacing the old note with a rest would have been the right thing do, imagine the reverse case - taking the quarter note C as you have it in measure 6, then hitting a button to switch it to voice 1. How would a program know you wanted it to come out like measure 2, as opposed to, say, actually putting the note in as a quarter note and deleting the sixteenths to make room for it? And note it would be wrong to simply replace the old C with a rest - that would leave you with unnecessary rests. This is a perfect example of what I mean by guesswork.

Anyhow, again, I can certainly imagine that occasionally a place might come up where a mistake was made, but those are easy enough to fix without needing to completely retype 10 measures as you described. I was more curious about the passage you mentioning importing via MusicXML that was somehow so messed up you had to redo it from scratch, but yet an automatic "move note to another voice, stealing time from the previous note and replacing old note with a rest" would have worked well. I still am skeptical that it would have been as simple as you say, and I think your example above demonstrates this very well. No single algorithm could change measure 2 into 6 but also change 6 into 2 - there would be two *different* algorithms necessary. That's what I mean about the program having to guess what you want, and being doubtful it would guess right often enough to be useful.

I'm not saying you are wrong; I'm saying right now, I still don't understand how an algorithm as you describe would actually pay off more than occasionally.

In reply to by Marc Sabatella

Like I said, it would notify you that there is a note conflict and do nothing. Most of these errors are not going to be having too many voices, it is having too few and having to transfer notes to a new voice. I don't understand what is so difficult about it, there is zero guesswork involved when you're transferring to an empty voice.

In reply to by darkNiGHTS

Marc, Sibelius has this feature. It works predictably. You might try downloading a demo version to see it for yourself. A specific use case for this (beyond minor corrections mentioned by darkNiGHTS) is writing four-part hymns on two staves, The use case is described in detail at… I wrote the article before I had ever used Sibelius, but Sibelius' implementation of switching single notes to a different voice is similar to the first proposal for non-modal multi-voicing mentioned in the article (although Sibelius' implementation doesn't work in note entry mode).

In reply to by David Bolton

Ah, the hymn example does make it a lot clearer to me. I hadn't considered cases where the rhythms were mostly similar between the two voices, because I almost never use multiple voices except to notate *independent* rhythms. Which is why the idea of simply moving a single note didn't make sense to me - it would have unintended side effects very often. I also managed to miss where Cory mentioned that he didn't need it to handle cases where you were moving to a non-empty voice.

OK, so I've got this noted now. I'm still struggling with understanding just what I can and cannot do within the plug-in framework, and actually am pretty sure I cannot currently access individual notes, but if/when that does become possible, I can try to see about handling this case that way. Not that couldn't be handled in the application itself too, of course, but I'm becoming increasingly interested in seeing what makes sense to do via plugin.

In reply to by Marc Sabatella

This is a real problem for me too. I need to import organ tunes from midi files and re-arrange as piano reductions. In its current form, Musescore is pretty much unusable for this simple task due to this single limitation. (I'm using version 1.1.)
It takes so long to delete notes, replace in a separate voice, then make all the unwanted rests invisible, especially as unwanted rests in the new voice may lie underneath a note in the original voice - it's just not worth the effort. I don't want to split all the chords into separate voices, just those occasional cases where the rhythm is not the same.

In reply to by rickfitz

Also, you might want to provide an example of a notation you are finding difficult to create - it may be there is a simpler way that you are using. Also, as noted elsewhere, it is usually incorrect to hide rests in a voice - a piece that actually uses multiple voices is supposed to show all beats in each voice. That's the way all notation style guides I know of recommend. The only common exception is that a single half rest in either the first or second half of a measure can be made invisible.

In reply to by Marc Sabatella

Marc, visible rests are recommended for instrumental music, but for homophonic choral works they are typically omitted. Because of the organ's close association with choral music and congregational song, it is common to hide rests for that instrument too (often the organist is actually reading from a four-part vocal score).

In reply to by David Bolton

Thanks to all for your replies. Hopefully the attached fragment will make my issue more clear. The first line is as imported from a midi file, the second line is how I need it to look.
So all I really need to do is change the tied notes into single full-value notes (as in the first bar of the attached file). I don't think this is really what voices are intended for, but I can't find any other way to do what I need.
An alternative approach would be to treat it as a choral score, with two voices per stave throughout. This would be fine, but would mean splitting every chord into separate voices, which is even more time consuming.
I have discovered that the unwanted rests in voice two can be deleted (instead of made invisible) which helps a bit.
Any other suggestions would be most welcome.

Attachment Size
example.mscz 2.34 KB

In reply to by rickfitz

Ok, this example is like the one mentioned previously - music that is mostly homophonic (all voices moving together) but in which every once in a while one voice adds an extra passing tone. Yes, this *is* what voices are intende for - this is clearly multi-voice muusic. It just happens to be simple enough and in a particular genre where the rests aren't needed. So It is true that this is another special case situation where showing rests is not needed. And that is why the facility to hide rests exists. It is important to realzie though that cases like this are the exception, not the rule, for multi-voice music. *Most* multi-voice situations are such that you *are* supposed to show the rests, and that's why the default is to do so. The alternative would be to not show them and thus complicate the more common situations where you do want them.

What would help in these special case exception situations, as I have suggested before, is a keyboard shortcut to hide the symbol just entered. What would help even more, I suppose, would be an option to set the default for rests in voices other than voice 1 to be invisible. It would also be nice if you could right click a rest, Select->More and choose all similar items in same voice, then right click and mark them invisible, but unfortunately this doesn't work for some reason - trying to apply properties to multiple rests at once only the affects the one actually clicked.

In reply to by Marc Sabatella

Thanks Marc. That's all very good and logical. If I were entering the score from scratch I'd use multiple voices, and your suggestion of an option to auto-hide rests would be very useful.
However, this doesn't help with the original issue that this thread is addressing, which is changing the voice of a note that's already in the score. It's a problem when editing a score that's been imported, in my case from a midi file, or in the original poster's case, a musicXML file.

Here's the current process: delete the note(s); switch to note entry mode and select voice-2; insert as many rests as are needed to reach the required position in the bar (measure); enter the replacement note(s); go back and delete or hide the unwanted rests. Then on to the next one... This feels very cumbersome, when all that's wanted is to move existing note(s) to another voice. It's ok for a few instances, but a real pain in a large score. The biggest nuisance is having to insert and then hide rests, just to be able to get to the required point in the bar. (Presumably there's no way to start a voice in the middle of an existing bar, otherwise it would be much easier.)

For a choral score, it would be necessary to change notes from every chord. An easier alternative might be to select all the notes and 'cut' them from voice-1, switch to voice-2, and 'paste' them back in. Sadly this didn't seem to work when I tried it.

As you say, imported homophonic music is a bit of a special case, but there are lots of useful midi files available for download, and a simple solution to separate out voices would be very useful. Maybe with a bit of practice I'll just get quicker at doing it the way it is now!

In reply to by Marc Sabatella

Was a solution to this issue ever created?

I deal with this issue almost daily:

I create a lead sheet of a song including chord symbols, lyrics, and melody in voice 1. Today's song has 28 measures and 8 verses.

When I am done engraving the lead sheet, I then create a guitar arrangement by adding a second voice for the accompaniment notes.

Only after engraving the lead sheet, I often decided that I would prefer the melody of the song to appear in voice 2 to create an arrangement in the style of Maybelle Carter with the accompaniment notes appearing in voice 1.

How can I switch all of the layer 1 notes to layer 2 and keep all of the attached chord symbols and lyrics?

In reply to by Marc Sabatella

Keyboard music is rife with hidden rests, particularly in contrapuntal music.

The attachment contains two fragments from Bach's Little Prelude & Fugue in G to illustrate this.

In the 1st example the music has become largely monophonic, and so rests are irrelevant, the two staves merely showing which hand plays which notes.

In the second example the redundant fugue entry in the tenor has had the following rest made invisible as it is a "false" entry, and putting a rest in there would add a fifth voice to a four part fugue!

Whilst entering this particular fugue I did come across the need to swap voices in a partial bar myself, having started the bar and then realising I was going to have to add Voice 3 in the RH stave. In this case I had reversed the two left hand voices for the first half bar for practical reasons, but realising I had to add a third voice meant they had to be put back so I could cross-staff the tenor. Fortunately only one bar was like this, but it was still an annoyance not being able to switch voices in half a bar.

In reply to by ChurchOrganist

Whilst being a novice w.r. to voices (for my purposes I find them a curse) I find that MuseScore 3 handles changing voice fairly well. Without having followed all the details above, my observation is that can only swap a voice if there is room in the intended recipient voice. If one voice has a crotchet to be moved to a part of the bar where it aligns with a quaver in the intended voice, it can't fit without other adjustments the user needs to make. Don't suppose this helps much, but it describes my limited experience of trying to get rid of voices. BTW, why is anyone still using MuseScore 1.0?

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