Triplets without brackets/numbers causing errors when opening file - MS3

• Dec 26, 2018 - 17:50

Hello, see attached. Here is the list of errors I am getting:

Measure 23, staff 1 incomplete. Expected: 2/4; Found: 10/16
Measure 66, staff 1 incomplete. Expected: 2/4; Found: 30/48
Measure 67, staff 1 incomplete. Expected: 2/4; Found: 30/48
Measure 68, staff 1 incomplete. Expected: 2/4; Found: 30/48
Measure 69, staff 1 incomplete. Expected: 2/4; Found: 30/48
Measure 70, staff 1 incomplete. Expected: 2/4; Found: 30/48
Measure 71, staff 1 incomplete. Expected: 2/4; Found: 12/16
Measure 75, staff 1 incomplete. Expected: 2/4; Found: 30/48
Measure 76, staff 1 incomplete. Expected: 2/4; Found: 30/48
Measure 83, staff 1 incomplete. Expected: 2/4; Found: 10/16
Measure 84, staff 1 incomplete. Expected: 2/4; Found: 15/24

It may be due to the fact that these measures are actually written as two quarter-note triplets, but the triplets are designated as having no bracket or number. The file is correct when I save it, but on reopening, several measures become corrupted, having the incorrect number of beats. I cannot simply delete the notes and re-enter the measure to correct — there is an extra 8th note that will not go away. The whole measure must be deleted and a new one inserted. There is nothing irregular in the measure properties dialog; the extra values are simply there.

Also, there is no way so select the triplet number/bracket after it has been turned off (saved file only, you can right after you've turned it off). It's as if the act of saving strips the triplet info from the file.

Also, please note the wide spacing of the sixteenth notes in measures 25, 85, 86 compared to the spacing in the measures next to them.

Measure 84 (which has an extra 8th note when the file is reopened), also has a lot of extra space after the barline (it occurs at the end of a line). There does not seem to be a way to correct this except by deleting the measure and re-inserting.

Also, a lot of adjustment was required to get the grace note slurs to look right when there is a second voice in the staff — a lot of collisions. Even without the second voice, when two sixteenth graces are slurred with a regular quarter note, the end of the slur collides with the notehead.


Attachment Size
Farewell_to_Stromness.mscz 41.16 KB


In reply to by Soolip

The slur spacing is a known effect of our algorithm, which first shapes the slur and then looks for collisions which it resolves by moving the slur further away. Would be better if we could shape the slur better in the first place if a collision would result otherwise. For now, this will just remain an area where manual adjustments may sometimes be helpful.

The score doesn't show a source, so it looks like this was made from scratch in version 3, is this correct? If it is, this is one of the biggest messes I've ever seen. All measures from 66-71 have an extra 8th note in them. There are other measures but this is where I focused to see what's happening.

Measure properties does not acknowledge the extra note and they are not identified as parts of triplets. I turned on all of the tuplet brackets and numbers so I could see them, but there are no tuplets being reported where they should in these measures. In measure 66, for example, the last 8th note is reported in the status bar as being on beat 1 of measure 67. The 16th note f in measure 67 correctly reports as being on beat one of measure 67.

This sounds like #280640: Attaching grace note to first note of triplet corrupts score . Most of your missing triplets include grace notes, but measure 72 is really messed. Just click the notes in order and notice how the beats of the notes are reported in the status bar.

P.S. much of this is to help people trying to fix tje problem, but any info about how you entered the score would help.

In reply to by mike320

It does indeed appear as if the graces before triplets are a problem — but it's not just that the the number that disappears, but the fact that there is a triplet at all. There does not seem to be a way to correct the issue, except by deleting the measures and re-entering them. Though they are corrupt again after saving and reopening.

In reply to by Soolip

It seems at the moment the only fix is to use version 2.3.2 for scores including grace notes on triplets. There is absolutely no workaround that would be acceptable short of lengthening the measure and making it look right - and that's not acceptable.

Note that the program manager has made this the highest priority to fix.

In reply to by mcx

Also, I've added a trivial repro file on, which appears to be the same issue. In my case I have both a grace note on a triplet (that lost it's '3' after the corruption) and also a beamless triplet (a quarter note and an eighth note). Not sure if that gives me the double whammy but the corruption is reported as

Measure 1, staff 1 incomplete. Expected: 3/4; Found: 21/24

In reply to by mike320

Hopefully staying in 3.0 and "leave out the grace notes until it's fixed" is a viable option. Seems like a high enough priority and I can be a bit patient. I might try squeezing it through MusicXML, but that hasn't been a good path for me in the past and "most lossless" is not confidence-inspiring (though likely accurate).

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