Display of bar repeats depending on layout and position (and other issues)

I would like to engrave Ravel’s Bolero in Dorico, and right from the start I’m running into several challenges. I think you’ll agree that there aren’t many pieces that need bar repeats as much as this one. So I thought I’d try to get a result like in the Durand score:

As you can see, the two-bar snare drum motif is repeated, the repetitions are counted up to the next rehearsal mark, after which the count resets. The motif is written out in full at the beginning of each staff. This continues for about 320 bars until the end. Of course, in the full score, there are no bar repeats; every note is written out.

So this is my first problem - how to achieve bar repeats in the parts layout but not in the full score? It seems to me that I would need a property for a bar repeat that, if selected, would replace the bar repeat symbol with the actual contents of the repeat, and that property would have to be accessible on a per-layout basis and for each individual repetition. As a bonus, an option like “Show full notation at the start of a staff” would be great. Such a property doesn’t exist - is there something planned, or is there another way I’m not seeing?

Second, is there a way to get the numbering as shown? For single-bar repeats, it works, but I don’t see how to do it with two-bar repeats.

Third, as you can see, there are no triplet indicators over the sixteenth-note triplets. I can turn them off (either individually or globally), but then there are thousands of signposts in the score that seem to slow things down (and I don’t really want to turn the numbers off globally because there are many other tuplets in the score that should of course keep their indicators).

I’m grateful for ideas :slight_smile:

Thanks,
Tim


P. S.: Another problem occurs right at the start when using the “Modern Orchestra” template - all the instrument names are displayed in English; setting the language to German in Engraving Options doesn’t change anything. From what I’ve read in the forums, that is intended behaviour, but why? What is the use of templates if they can’t be used in other languages than English? The instrument names themselves (in setup mode) are partially translated, but wrongly (as I’ve mentioned years ago: E flat is “Es” in German, not “E Be” - this error is present in all instrument names that have a “flat key transposition”).

P. P. S.: In the Durand full score, the staff size is progressively reduced during the piece, in order to fit more and more instruments onto the page. I don’t think anything like that is possible in Dorico either, is it?

Your main question and your P.S. are outside of my comfort zone. Your P.P.S. certainly can be achieved in Dorico. Select any Frame Break or System Break (or add one), and from the bottom properties panel you can change the space size. You can then set it to something different at the next System/Frame Break (or any other System/Frame Break).

There is no way to show a bar repeat in the part but the music written out in full in the score, though that is something we would like to do at some point. It is also not possible for Dorico to automatically show the music at the start of the system, and that is unlikely to ever be solvable automatically because of the circular dependency in the problem: Dorico can only know when a bar is at the start of the system after the music has been cast off, but it cannot cast off the system until it knows how many bars will fit, but how many bars will fit depends on how much music must appear in those bars, the answer to which question is different based on whether the bars show actual music or bar repeats. The only way this could conceivably work would be for Dorico to determine that it should show bar repeats automatically, but it would not be able to cast off the music differently as a consequence, so you wouldn’t actually get any more bars on the system this way, which eliminates at least some of the benefits of using bar repeats in the first place.

You’ll need to manage the numbering of the bar repeats manually, too: Dorico only numbers successive one-bar repeats automatically at the moment.

Thanks Daniel and pianoleo!

I understand the problem with circular dependencies, manually setting a flag for “written-out” should be fine when than functionality is added at some point. For now, I think I will copy the actual music so it looks good in the full score - and maybe create a second snare drum player who gets to see the bar rests.

For the triplets, I probably need to do something with hidden time signatures as described in this blog post: https://www.scoringnotes.com/tutorials/masking-meters-creating-polymeters-metric-modulations-dorico/ - I hope this will also work with cues and can be changed back and forth (since many players switch between 3/4 and 18/16 feel).

Apologies for revisiting an old thread, but I’m wondering if any progress is being made regarding bar repeats on score vs. parts. As I learn more and more about Dorico, I find that virtually every feature has been carefully thought out and implemented with consideration for practical application – EXCEPT for the bar repeats feature, which in its current state seems not to be useful in most situations. It’s typical for the notated repeated measures to appear at the start of each system. This means that unless the score and part have the exact same system layout, there is no way to use bar repeats without some time-intensive workaround.

Any chance this is a priority?

We understand the requirement to be able to show the music to be repeated at the start of the system in the parts, and that you might also sometimes want the music to be written out differently in the score (e.g. you might not want to use bar repeats at all in that context) but unfortunately at the moment this is problematic to achieve. We certainly anticipate returning to this area in the future, but I’m afraid I can’t say for sure when that will be.

It’s been 7 months - so an old thread… but I need to ask if there has been any movement on the option to display bar repeats in parts, but the actual notes in the score?

My current workaround is to create a double set of players - one for parts printing, and one for creating a full and/or condensed score.

Is there a better/quicker/easier way to get bar repeats displaying in parts and not in the score?

Not yet.

I’m trying to imagine how Dorico could automatically show the first repeat in a system, but Daniel has a point: the problem is circular. But I don’t like the idea of casting off as though the music were there.

So I’d propose an assisted manual solution, where the user can select arbitrary bars in a repeat region to write in full, customizable per layout. That way, you’re not even limited to the beginning of a system. But a plugin could help facilitate that typical use case.

The plugin would simply go through the score system by system, looking at the first measure in all staves. If there’s a repeat region, toggle the “show music” property, insert a system break, and continue, jumping to wherever the next system starts now.

So when I’m done writing music, I could select all repeat regions in the score and reveal everything, then go into each part that contains repeat regions and run the plugin, tweaking where necessary.

1 Like

Looking back at the Boléro example, Durand’s engraver obviously had some way of determining how to cast off this part (or any other part, for that matter). But if any skillful engraver ever was able to use this method, why can’t Dorico?
I can’t help it, but my intuition tells me it’s incorrect to say the problem is unsolvable because of circularities. If that were true, no copperplate engraver would have ever been able to solve it either. Yet they did, routinely. Ergo: there exists a finite algorithm. No doubt there’s repetition and/or recursion involved, but that’s not necessarily circular. There will be end conditions like ‘doesn’t improve anymore’ or ‘good enough’.
Being a software developer myself, I’m quite aware translating a set of intuitive rules-of-thumb into software can be daunting task.

I think Dorico casts off by measuring the required spacing columns in each measure, then calculating how many systems are necessary. This would likely be O(n), where n is the number of measures. As Daniel mentioned, the issue is that a repeat-bar will have two different measurements for how many spacing columns it will require, and Dorico can only know that a bar begins a system after casting off.

So after casting off once, I suppose Dorico can look for the repeat-bar measures that needs expansion, and cast off again, using the new spacing column measurement. But it can only expand the first repeat-bar that begins a system, as the measures will likely have moved after expansion. The process could continue, casting off again, every time that a repeat-bar measure is encountered at the beginning of the system.

Assuming that the expansion of a repeat bar measure does not affect the layout of preceding systems (which may not be the case), this algorithm should terminate, but would take O(mn), where m is the number of expanded repeat-bar measures.

If such expansion does affect the layout of preceding systems, then the casting off might oscillate, causing the expanded measure to get pushed back to the previous system, and therefore no longer need expansion, and then it would return to the beginning of the system… So Dorico would need to keep track of each casting off, adding internal system breaks to ensure a state does not get repeated.

Of course, I don’t really know how Dorico works under the hood, and my assumptions may be incorrect, or ignorant of a bunch of other considerations that would make this feature prohibitively difficult…

Speaking as a non-programmer, perhaps Dorico could cast off parts in situations with bar repeats in a similar way to its treatment of condensing, which one would conceivably apply only at the end of the formatting process and which could then avoid the problems of circularities.

I solved the Boléro tambour problem as follows:

  • I created a new flow, just for the tambour (= a “fake” flow)
  • I noted the bars to be played and repeated, and added manually the number of times to be repeated, bar change numbers and the same rehearsal points as in the Durand edition.
    The result is a one-page score, instead of 7 unusable pages.
    Here is the result:

(By the way: the tempo indication in the Durand edition is 72 bpm, but there is evidence that Ravel in fact wanted it slower: Ravel told in an interview that the work (of 340 bars in 3/4 beats) takes 17 minutes. That results in a tempo of 340 × 3 / 17 = 60 bpm. See: Calvocoressi, M. D. (11 July 1931). "M. Ravel discusses his own work: The Boléro explained". Daily Telegraph. reprinted in (Ravel & Orenstein, A Ravel Reader: Correspondence, Articles, Interviews, 2003, p. 477)

2 Likes

I know it’s two months later, but you linked here from another thread and I got thoroughly nerd-sniped. An issue with your solution is that “repeat 50 times” doesn’t reflect the musical structure at all and doesn’t provide any feedback to the player whether or not they’re still counting correctly. Also, what if the conductor calls out “start from rehearsal 13”?

In that light, here’s my solution. Granted, the MusGlyphs dynamic indications are a bit of a stunt solution, but it saves on writing out the pattern again. The casting off reflects the structure, which is AABB-AABB-AABB-AABB-AB-Coda, and also the cue texts (arguably the most important markers here) are straight from the original Durand part.

1 Like

Dear Hugo,

Your suggestion is brilliant, thank you! The idea of using the eight rest in bars 147, 165, … when the dynamics increase is really clear for the player.

We are planning to publish our Boléro’s version, and if you agree, I should like to use your idea for the tambour score. We shall happily give you the credits for that! (You can easily find me on LinkedIn.)

By the way: I see that in bar 339, you write what was published in the Durand edition. except that you add an eight rest for one of the two tambours. I suspect the Durand edition may contain an error, because the score indicates that they start playing “à 2” as from bar 291, and I see no musical reason why the first tambour would not play the first eight note of bar 339 (and Durand doesn’t write the eight rest). Particularly, when looking at Ravel’s own transcriptions for 4 hands on one and 2 pianos, the first beat of bar 339 is flayed fff. So, why would the tambours not play that beat in the orchestra as loud as the previous beat?

Robrecht

Sure you can use this approach, credit not neccessary but always appreciated :slight_smile: On second thought I’d also parenthesize the rests in front of the dynamics so it’s even clearer that those aren’t literal rests.
About bar 339, it’s notated this way (with the rest) in the original Durand part, but indeed not in the score. Maybe it’s meant to be some kind of two-stick hit but there’s no explanation.