Voltas without repeat bars

• Sep 28, 2020 - 15:45

The attached image shows a general layout of a waltz score I am copying. This represents an entire section ending with section break. Starts with several measures collectively labeled A followed by alternate endings B and C. C is labeled Fine. This is followed by a repeat consisting collectively of measures D, with endings E and F. Second ending F is DC al Fine. If I set Play Repeats in the DC the measures are played back as A, B, D, E, D, F, A, C and exit to next section. This is expected. If I uncheck play repeats I get playback of A, B, D, E, D, F, A, B, D, F. In this case the jump back to the beginning does not result in the playing of the second volta ahead of the repeat (I thought second endings are always played in jumps) and when the repeat endings are encountered the second ending is selected without any repeats and exit. I am not sure of why the counts get reset on the first set of voltas in the DC pass as if it is first time through, or why the second volta is not chosen.

Attachment Size
DelCapo.jpg 43.44 KB


The repeat count of a volta is always compared against the normal repeat logic; meaning it only takes barlines into account and not jumps. Jumping in itself does not influence the possible repeat count.

So what happens after you jump is taken is that it jumps to the start of the score and plays back the "final/last" repeat for that measure. The start of a score/section works as an implicit start-repeat barline. Looking from there on, it has no matching end-repeat barlines. This results in the last repeat count being evaluated as repeat 1.
Thus playback is started again from the first measure using repeat 1. Upon hitting the 1st volta, it is honored. Then the 2nd volta is skipped because it doesn't match the current repeat number.
When starting from D, the start-repeat is encountered and it's "last/final" repeat is determined. It has one matching end repeat barline encompassing a single repeat instruction. The last repeat is thus being evaluated as repeat 2 for that start barline. The 1st volta of E is then skipped, as it doesn't match, and the one from F is honored.
At that moment the Jump has already been executed, so it is now ignored and we've run out of score, ending the playback.

A more common way to write this out and which works is to:

  1. Replace the D.C. al Fine with a D.C. al Coda
  2. Remove the 1st volta from B
  3. Add a "To Coda" marking in the final measure of A
  4. Remove the 2nd volta from C
  5. Move C to the end of the score
  6. Place a coda at the start of C


Attachment Size
311041-dcAlCoda.mscz 6.74 KB

In reply to by jeetee

I need to read this slowly in detail. I notice your modification to use a coda. So after the Coda (C) is played, the score starts on the next section. In the original Fine placement, the jump is default and score continues with the next measure after F which has a section break. Do I add a section break to the end of C as you drew it up and place following section measurers after the C section or do I need to mark the beginning of the next measure with a Segno and give it a unique name and provide a jump to it somehow.

This issue appears to be one of not properly marking repeats, there being no end repeat marking. Suppose I added a normal end repeat to separate B/C. By default this would play twice which I do not want. Can you mark the play count to 1? Would the counting work on the DC pass when the measure is encountered for the second time?

In reply to by msokol

Please note my original post that when play repeats was chosen the the playback worked okay for whatever. It was just my trying to understand why NO PLAY REPEATS resulted in failure. IT was the attempt to understand Play/No Play repeat options and what the affect on the play back was for the different options.

In reply to by msokol

I can not confirm the behavior you've described in the original post with play repeats enabled in 3.5.0. See attached example score, which should match your original image.

To further answer on the previous post; if after the coda you want to start a new section, then yes, you should insert a section break at the end of the coda.

Attachment Size
311041-dcAlFine.mscz 6.44 KB

In reply to by jeetee

The attached score is the one that has the behavior described. See section labeled as No. 3. The DC al Fine is set to Play Repeats. It plays the first volta skipping the second marked fine. It then plays the main simple repeat with alternate endings and jumps back to the beginning of the section and plays skipping the first volta, playing the second and then exiting the section to the one following. I believe this is the pattern a provided in my schematic.

Attachment Size
Strandbilder.mscz 50.56 KB

In reply to by msokol

As I half suspected, the reason that it "works" currently is a bug.
In 3.5.1 it'll no longer work, just like it doesn't work when you have no other section in front of it.

The bug is that to determine that "last" repeat after a jump, the current version goes looking backwards through the entire score for a start repeat reference (and section start didn't count as one).
In this case it "finds" the start-repeat of m70 (= marked as m17 of the previous section), which has a matching end barline at the last measure of that previous section, leading to the repeat count of 2.

Starting from 3.5.1 a section start is also treated as a repeat-start reference point (as it should be) which in your case will result in the same playback in the full score as in the isolated test case.

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