musescore inserts extra beats in measures

• Sep 18, 2011 - 21:17

I posted this question previously, but did not include enough info. I apologize to the community.
and now I have more information.

Musescore 1.1, Windows XP SP3
I entered a melody line using triplets. Then I copied the melody line to the instruments playing harmony, then changed the pitch of the notes. No problem. But when I tried to copy staves both horizontally and vertically (several measures and several instruments) to paste farther along in the piece, musescore dropped the triplet designation and left me with quarter notes, and six beats in a measure instead of four. I tried deleting the notes, deleting the measures, copying over them--all sorts of things--and when I paste a valid measure into one of these aberrant ones, it adds two beats to the following measure. I captured a screen shot of a couple of examples., attached.

It's like a virus, each time I correct an infected measure, the following measure gets infected.

Thank you.


Did you copy a partial measure? That's the number one recipe for creating corrupt measures - copying a partial measure containing triplets, eecially if you copy it to a different position within the destination measure. But more helpful than a score that shows the corruption would be a score that doesn't, and specific instructions on what to copy and where to paste it to reproduce the problem.

I've had a similar problem: I wrote a score last week involving triplet crotchets, saved it and closed it. Then this morning I opened it up, and MuseScore had "lost" some of the triplet markers and had spread the crotchets out over the bar. But since there are now too many notes for the bar, it appears to have stuffed the "extra" notes into a floating half-bar after the full bar. At any rate, in several places the lines are now out of synch.

EDIT: Just to add, I did not copy/paste any partial bars.

Cheers, Mark

In reply to by markgcarroll

I copied a small portion of the score. All three measures are supposed to have the same triplet rhythm, but musescore dropped the triplet designation in the middle measure, thus adding beats. When I try to copy the correct (first) measures over the incorrect (second) measures, musescore adds beats.
musescore 1.1, windows xp


Attachment Size
problem with added beats.mscz 2.64 KB

In reply to by trustin

I don't see any added beat in the second measure. I tried to copy paste the first measure to the second one. I also tried different copy paste combinaison without having any issue. Do you really see a problem in measure 2? Can you give exact step for copy pasting (I use Ctrl + C and Ctrl + V, to select I click on a note or a measure, and shift click on the last one)

In reply to by [DELETED] 5

The alteration occurred between closing the file and opening it up again. I entered the following rhythm in one 2/2 bar: triplet crotchets (rest, crotchet, crotchet), dotted crotchet, quaver. Then in the following bar I entered the same triplet crotchet arrangement, followed by two lots of sextuplet semiquavers. Unfortunately I can't remember whether I copied and pasted the triplet crotchet arrangement in the second bar or manually entered it.

Hope that helps! :)

In reply to by [DELETED] 5

Can MS help with reproducing these sorts of steps? Sometimes fixing an issue is not nearly as difficult as the process of tracking down and isolating the reproduction recipe. Am wondering whether there is information in the MS "session" that could help list out the prior steps? Or maybe there is some other Qt "recorder" tool that could capture the steps? I've run into quite a few MS oddities, but they seem to always happen when I'm concentrating on the music itself and not necessarily paying attention to every editing step and so are difficult to isolate after the fact.

In reply to by mtherieau

It's possible to build a "debug" version of MuseScore, and maybe there would be a way to coax some sort of log from that, if one were using it when one noticed the problem appear. No idea if that would prove helpful, but if so, I'm happy to install a deug version and save any logs while I play around with triplets and try to reporduce this.

While I don't have a specific series of steps yet, this happened three times in the last day (during which I was actively using MuseScore almost the whole time I was awake). I have been careful not to copy/paste partial measures involving triplets, but here are the scenarios in which problems occurred:

1. A measure that was created using quarter note triplets with a rest on the first beat, then the measure was copied to another staff (at the same time position) then the rest was replaced by a note, and then the last note of the previous measure was tied into to the beginning of the measure with the triplets. At some point later after program crash followed by a reload and a "restore previous session", the triplets turned into quarter notes, with a corrupt measure.

2. At some point after creating the above but before the reload that caused the problem to show up, I copied measures adjacent to the one I started with to a different measure later in the score but in the same staff. The same reload that exposed the first problem exposed a similar one here.

3. At a totally different place in the score, a guitar part was notated with two voices, no triplets in sight. I later edited that part, adding a third voice, changing the time values of and otherwise editing some of the existing notes, and used edit. Oe to move some of the rests to more visible locations. Copy and paste of these measures on top of existing measures produced an extra quarter rest in one of the bars.

In all three cases, I was able to recover by deleting the problem measures completely, inserting new ones, and re-entering the original contents. Actually, I chose to live dangerously - I copied and pasted as much as I could out of the old corrupt measures, as the corruption seemed to only affect one staff.

In reply to by Marc Sabatella

not sure about tuplets, but multi-voice paste has some known issues. One of those issues is:
Undoing multi-voice paste does not produce multi-voice rest on previously occupied note

The root cause of this issue seems to be that the paste operation does not take care to remove the secondary voice elements before pasting in the clipboard contents. As noted that affects undo history, but it may also have further consequences too. To avoid problems, it may be a good idea to clear the contents of the destination staff before pasting a multi-voice selection.

somehow, musescore added an eighth note to this measure. I tried selecting and deleting the measure, but musescore put the extra eighth note into the next measure.

I copied some of the score to a blank staff. The first measrue is OK. The second measure looks OK but is corrupted. The third measure is the one that the extra half-beat was inserted into. The fourth measure is what was on the score, with the extra eighth note. If you copy the fourth measure Ctl-Z and paste it into the fifth measure, you'll see what I mean (I hope)


I've tried for some time to duplicate this issue, and of course when you want to you can't. Most would agree though that the easiest way to create corrupted measures is to copy and paste some partial of any tuplet as it creates the most hideous mixture of notes/rests that no longer add up to the correct beat count. I suspect this is how measures following this incomplete measure get corrupted. It is very easy to accidentally select a partial tuplet, either its first or last note, and thus starting you on a downward path to a very corrupt score.

Should this not be dealt with first, either by disallowing the copy of a partial tuplet, or by automatically extending the selecting to include the entire tuplet, or by filling in the missing tuplet notes with tuplet rests to complete the beat?

In reply to by schepers

I would agree that pasting an actual partial triplet is going to be frought with problems and might not mind if it were disallowed. But I don't know that I've ever actually done that. What's more common, and unfortunately also problematic, is copying, say, the first two beats of a measure when one of those beats is a complete eighth note triplet. Actually, it's quarter. Ote teiplets that seems to trigger problems more often, but the same applies - I mean, copying two beats that contain a full quarter note triplet.

And actually, I can think of one case of copying a partial triplet that one would expect to work, and that's if pasting it onto an existing triplet. Pasting the second two thirds of one triplet onto the second two thirds - or even the first two thirds - of another teiplet seems like something with half a chance of working. Certainly, copying one note from a triplet onto a single note of another triplet might be expected to work.

Of course, while it would be very cool if it turned out one single bug were responsible for all the problems of this nature that have been encounteed, I fear it will turn out to be not so simple. But it might well turn out to be one incorrect set of assumptions or ways of handling notes that is used several places in the code. Wish I were familiar enough with the code to be able to investigate stuff like this, but I think this is the kind of thing someone like me would be likely to make worse.

Yes, I have pretty much the same problem, and it also affected rests. if I select a whole measure of "bugged" rests and press delete it looks okay again but the following measure gets an extra eighth note. I kept on doing that for a while and it simply went crazy, check the attachment! Seems like the easiest way to prevent this is to copy one instrument at a time.

Attachment Size
Musescore bug.JPG 128.41 KB

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