Weirdness with Start End Repeat Bars and unusual Time signatures

[pmc.galoubet]

The piece is a Renaissance Christmas song, with a couplets section, a refrain section and a ritournelle, and will be repeated a number of times. The piece has basically bars the length of which is 4 half notes. The melody staff will have a Time Signature of 4/2, but the percussion will have [3+3+3+2]/4. Both signatures align perfectly. But the start of the couplets is longer, so the first bar will actually have a Time signature of 5/2, with [3+3+3+2+2]/4 for the percussion. Thus, each staff needs to be given an initial signature for one bar of extra length, then the shorter one for the rest of the piece. There is no pickup. There will be no need for any other Time Signature Change in the score. The melody has an extra staff for ossia stuff, All Barlines will go from top to bottom. The first 2 bars need to be sung twice, so need to be enclosed in a repeat structure, so for playback there will be 2 extra Time Signature Changes.

NautreiSiamEiciVenguDZ.dorico (1.8 MB)

The project as it is looks perfect when printed (without signposts!), but I already had to cheat to get that result, and it does not really play back perfectly.

First problem is that the piece needs 20 bars, and I could not get rid of an empty extra bar. I Moved it to the 2^nd page, which I will not print (quite sloppy).

The second problem is that I had to add extra Time Signature changes on bar 2, which I do not understand the need of, and that I hid. If I delete them, all the following stricture is broken rhythmicwise.

Probably related to the previous problem is that the repeat structure is ignored in playback.

And of course, proofreading has a lot to say about it:

Several times. Both clear and unclear. Such Repeat Bars are not created by one action. During my fight with Time Signatures, I occasionally saw an unwanted Repeat Bar attribute appear, which I had to remove, but what makes the final structure unacceptable?

Probably related to the previous one.

I know I did that one. But I do not really understand it. If this should not be repeated, why do I have to repeat it if I want to keep my structure?

Twice, unfortunately unspecified bar. I am totally clueless. There are exactly 2 lengths used here, whether in time signatures or in actual bar content: 10 quarters and 8 quarters. And since there are no pickups, there are no “,d” adjustments in time signatures.

A lot. At least they are not a surprise.

I have the feeling that the whole structure of the first system is the root of the problem, but I do not undestand why.

[asherber]

@pmc.galoubet

First problem is that the piece needs 20 bars, and I could not get rid of an empty extra bar. I Moved it to the 2nd page, which I will not print (quite sloppy).

Select the frame break on page 2, and delete it. Then open the Bars popover with Shift+B, type trim, and hit Enter.

@pmc.galoubet

The second problem is that I had to add extra Time Signature changes on bar 2, which I do not understand the need of, and that I hid. If I delete them, all the following stricture is broken rhythmicwise.

I don’t see a time sig change in bar 2, but I do see them in bars 13 and 14. I first deleted the one in bar 13, then the two in bar 14. This made another change (cut time) in bar 14 appear, which I also deleted. I think that leaves all the bars with the correct number of beats.

There are a couple of warnings about repeated time signatures in the Galoubet additional staff. If you slide the +1 staff signposts to the left at these spots, you’ll see the time sigs, and you can delete them (and then slide the signposts back).

This leaves 4 warnings about the repeats. These are probably related to the independent time signatures, but I’m not sure. Maybe someone else will have an idea about how to fix those. (There’s nothing wrong with just ignoring them, but as you noted, the piece currently won’t play back correctly.)

NautreiSiamEiciVenguDZ edited.dorico (1.8 MB)

[johnkprice]

This version eliminates all unnecessary independent time signatures and plays the repeat:
NautreiSiamEiciVenguDZ-revised.dorico (1.8 MB)

There is one benign proofreading issue, and the bar numbers have changed because the engraving option Bar numbering for repeated sections is set to Count repeats.

[pmc.galoubet]

Hi John, nice to hear from you again !

I was preparing to answer your first sendings, to thank you for cleaning up the sloppy stuff I had left and leaving a clean single problem to solve! But you were faster on the draw. How did you fix it?

[johnkprice]

On a copy of your file in galley view, I deleted the -1 staff signposts and added a +1 staff signpost at the beginning so the added staff for the galoubet was visible throughout. I observed that there were time signatures at the start of bars 1, 2, 3, 13, 14 and 21. I also observed that there were barlines changes at the start of bars 1, 3 and 21.

Then I opened another copy of your file in galley view and added +1 staff signposts to the start of bars 1, 2, 13 and 20. I deleted the frame break, time signatures and final barline at the start of bar 21 and trimmed the flow. I deleted every remaining time signature and the repeat barlines. I added global 5/2 and 4/2 time signatures at the start of bars 1 and 2, respectively. I added the start repeat barline, changed the time signatures for the tambourin to independent 3+3+2+2/4 and 3+3+2/4 signatures, and added the end repeat barline. Finally, I deleted the four +1 staff signposts I added previously.

[pmc.galoubet]

Everything is quite clear to me except for the repeat structure problem.

First I will say that as such the project is both “picture-perfect” and “playback-perfect”, the issue is solbved and I maquite happy, thanks to you. But I remain confused about the above problem, and I would like to have your opinion.

I opened your revision, and unhid the Time Signature Changes on Barline 3:

while in my version this was

Experimentally I tried to change the tambourin Signature to [3+3+2]/4, and immediately, the repeat ceased to work, and the proffreading agqin compained:

So this is the main fix for the repeat proble, isn’t it?

From and editorial POV I would not like to see anything but 3+3+2 on the tambourin, but since it can be hidden, this is not a real problem. Bu it is notclear to me why this is the needed fix, even if it is cleared “related to the independent time signatures” like you said initially.

But more than that I still do not undestand whi Time Signature Changes are neccessary on Bar 3. The Time Signatures st on Bar2 should be persistent past the Barline 3 (which also happens to be the End Repat Bar!). This suggests Repat Pairs work something like a transactional Begin End block in a progamming language, where everystate change made inside would be restored after the End. But as far as I know there is no such rule in score wrting? So, is that the explanation of the behavior, and if yes, is it the expected behavior or an artefact of mixing too closely independent time signatures and repeat structure?

What is more, is this really the explanation, why would the tambourin timechange have to lose its independance?

I am not sure if I make much sense, what do you think?

And thanks again for your help anyway.

Patrick

[johnkprice]

Whenever a barline is changed in a Dorico project, a (usually hidden) time signature is also added. In my version of your project, the end repeat barline has a hidden global 4/2 time signature. Why does the repeat stop working if the tambourin is given an independent time signature of 3+3+2/4 at the start of bar 3? Honestly, I don’t know. I tried several possibilities before posting my version and concluded that a benign proofreading issue was preferable to keeping the repeat from working.

[pmc.galoubet]

Absolutely preferable yes !

I atill would like to understand it better, and what you tell me helps a lot, that changing a BarLine from normal to Repeat Bar will automagically generates a Hidden Time Signature (How did you learn that?)

I guess that behind the scenes, a Repeat Structure does work like the transactional Begin End block I mentkioned, and to compensate dorico generates a copy of any inner changes. Now I know that duplicate looking time signature are necessary in that case, the only remaining problem is why did Dorico generate Galoubet: 4/2, Tambourin 4/2 and not 3+3+2/4. Maybe because it doesn’t matter, they will be hidden anyway. And the reason for not being able to “fix” it (which would not be necessaety anyway) may not be the value change, but the fact that I am adding an extra level of Time Signature that cannot really take the place of the doco-generated one.

Maybe I have understood everything (fingers crossed)

[johnkprice]

@pmc.galoubet

changing a BarLine from normal to Repeat Bar will automagically generates a Hidden Time Signature (How did you learn that?)

See the Note in this help topic.

[pmc.galoubet]

It would be a bonus to have read every line of the help file … and to remember it And to update oneself when dorico does!