Another 3.5.1 divisi condensing bug

Bars 170-180 in flow 5 (may be similar issues elsewhere that I haven’t noticed), the 1st violas disappear no matter how I condense them (fine if I force them to be uncondensed). I had this problem before and it turned out to be because one of my divisi parts had been input in a second voice. That’s not the case here. Am I missing something?

Also I notice the staff label over the page doesn’t include 1.2. I’m pretty sure both these things were working correctly in 3.5.

http://www.prestopublications.co.uk/condensingbug.zip

I’ve found another example, this time in the horns (i.e. solo players, not section divisi) at bar 225. Try condensing horns 1-4 into a single voice (they’re in unison). For some reason the first is removed. I can just change the label here (from 2.3.4 a4 to just a4) but maybe this will help identify the underlying problem?

Can you please attach a screenshot of what you’re seeing in each relevant place? This project is so big that although it’s crucial that we have the file (for which, thank you), but in these cases it would also be helpful to have screenshots of what you’re seeing so that I can more easily orientate myself.

No probs, see attached


I might add that there seems to be something funny going on with bar numbers (hence calling the viola image before fig 41!) - they seem to be different between write and engrave modes, but I haven’t looked at this in detail - I might have done something silly…

Thanks very much for the pictures. I’ve spent a while looking into both of these issues. Unfortunately it’s quite impractical to debug projects as large as this, so I have spent a while trying to cut the project down into two smaller projects that still reproduce the problem.

I’ve managed to narrow the viola case down to the presence of the preceding condensing change that applies to the viola some 50 or so bars before, and removing that condensing change will allow the second divisi section to appear correctly; it may be that adding a further condensing change that resets the viola condensing back to its automatic default will resove that. We will nevertheless look into it further to see why there is a change in appearance between Dorico 3.5 and Dorico 3.5.10.

In the case of the horns, unfortunately simply opening the project you attached doesn’t give me the same result at bar 225; the horns are not condensed at all there, which I put down to the presence of the local time signatures at bar 227. I suspect that will be the root cause of the instability in the result there.

Brill, thanks Daniel. Will mess around with extra condensing changes and see if I can get it all to behave.

Oh, and I can’t get to the bottom of those bar number oddities - do they look wrong to you too?

To be honest I didn’t look in detail at that, but if you’re using local time signatures, that will be the root cause.

Hm, I can understand how they might get out of sync if I have different concurrent time signatures, but up to this point I don’t have that anywhere where the number of beats in a bar doesn’t add up (i.e. it’s mostly 3/4 against 9/8 or 4/4 against 12/8, with the first bar of the compound time signatures an anacrusis of 6 and 8 quavers respectively, with hidden tuplets and a hidden returning time signature in the following bar). And even if I had missed one or two, surely write mode should match with one of them, and surely the gaps wouldn’t be so large?

For example, in the same file the bar numbers go 235,236,794,795! This is the same in write mode in page view, but shows 237,238 etc. in galley view. N.B. From 794 they’re actually (counterintuitively) correct, as I have a bar number change at the very beginning of the flow. Perhaps galley view always shows the bar numbers from 1 and I haven’t noticed, but either way the page view and engrave mode are definitely wrong…

I’ve got the condensing problems pictured above to behave:

The horn one was easy - you were right about the independent time signatures, somehow I failed to notice that one of the 4 horns was missing its 12/8 and hidden 4/4.

As for the violas, no combination of condensing changes seemed to make a difference to me, but as I was preoccupied with the subject of independent time sigs I plonked a hidden 3/4 into both viola parts at the beginning of the broken bar (even though both were already in 3/4) and lo and behold the first viola appeared. Hope this is useful to somebody!