Bug Report: Key signature changes in SATB condensing score

When a key signature change occurs mid-line on an SATB condensing staff, the bass clef signature appears as-if on treble clef.
To reproduce…

  • Add SATB instruments
  • Set up S+A and T+B condensing
  • Add a time sig and insert some bars
  • Add a key change in bar 2
  • View in condensing mode

The sharp/flat signs on the bass clef appear in the positions they would if it were a treble clef. eg for G key sig, the # appears on the top ‘A’ line, rather than the second F line.

Presumably it’s looking at the first instrument in the condensing group’s clef when deciding where to place the #, rather than the actual clef in use. Indeed, if I change the tenor part to be in bass clef, they jump to the correct positions, and if I change it to tenor clef they jump to their tenor clef positions.
clef bug.dorico (474.2 KB)

Yes, this is a known limitation of using custom condensing groups that use different clefs. When we return to do some more work specifically on condensing of choral music, we’ll sort this out properly.


Please at least consider a ‘quick fix’ to use the same algorithm that you use to position the key sig at the start of a line to position one for a key change.

As it is, this ‘limitation’ (don’t use condensing on staffs with different clefs and key changes) pretty much rules out vast swathes of choral music.

I know there are lots of things you want to do to make SATB condensing better, but most of those are presumably things to improve the workflow and to make more things possible - this doesn’t feel like that, as it is silently producing incorrect output.

I only noticed the error while actually singing in a rehearsal with 40 copies in the hands of the rest of the choir - my proof-reading didn’t extend to checking that the software was working correctly!

The team has been vocal about the limitations of condensing vocal music since the feature was first introduced some time ago, and this isn’t the first thread about it either.

The way to deal with it for now is to just create separate players where you combine things manually at those spots where you need it. Two generic “voice” instruments renamed SA & TB (or simply using the “choir reduction” grand staff instrument) takes care of the problem nicely.


Yes, I’d concur: Condensing’s main benefit (or point) is to consolidate material on one staff in one layout that needs to be displayed separately in another layout (e.g. parts, score).

Assuming that you’re not giving individual parts to your singers :astonished:, just use different staves, two voices per staff, staff hiding, divisi, etc to achieve any changes in staff topography, if necessary.


Hell from playing piano pieces where clefs can appear anywhere I’d hardly notice - worst was some Impressionistic piece I sight read yesterday where left was treble and right was bass. That double cross was a first (for me), and made for a full stop, usually they’ll be a little more kindly and just have one hand cross the other.

Anyhow pretty minor problem IMO FWIW unless you’re publishing a book or something, took me staring at your example to figure out the problem, I just notice the shape (number) of it anyhow on bigger scores.

1 Like

Thanks for all the ‘helpful’ advice everyone.

I’ve found a workaround, which is to force the tenor part to bass clef around the key change.

To answer some of the comments above…

Yes, I do need all the parts separate, not for printing, but because the score also needs to be used to produce ‘stems’ for each part individually for mixing down into rehearsal recordings. That does mean, however that I’m free to make the ‘parts’ as ugly as required to suit the condensing as long as the notes are correct.

I am aware that the team have been vocal about the limitations, but I have just used it to condense down an SSAATTBB + solo & piano score into a mix of S-A-T-B, and SA-TB with everything looking exactly as I wanted it. Yes, there were times when I have to use workarounds to get it looking perfect (eg deleting/moving lyrics on some staffs, and hiding a myriad ‘a2’ markings), but honestly I found it extremely good, far easier than I feared from all the warnings. I think the only compromise I had to make in the end was not having join/split arrows when the condensing changed, and having to hide part names at the start of staves. Even those I probably could have got working if I’d been prepared to spend a bit more time in Engrave mode.

This issue was the only one that actually produced incorrect (rather than aesthetically sub-optimal) output, and the only one for which I couldn’t find a workaround (until now).

I’m not sure that ‘I’d hardly notice’ is a reason to ignore it. I didn’t notice either for quite a while, until I happened to flick my eyes back to the key sig during rehearsal and got confused as to what clef I was in and started singing wrong notes. It’s amazing how something so minor can completely befuddle your brain in the moment. I do intend to publish… only to CPDL admittedly, but would be nice to have it right.

I spent a long time trying to word the request for a fix so as not to make it sound like a criticism - as you can see above, I find it extremely usable. Maybe I should have said ‘fixing this would open up vast swathes more choral music’, rather than saying that it ‘rules out vast swathes’.

I’m eagerly awaiting the updates to make things even easier, but in the meantime will continue to ignore your advice to not bother condensing vocal music - it works just fine thanks.

For anyone interested, the piece was CV Stanford’s ‘Songs of the Sea’ with SATB chorus. The only versions out there seem to be full version with TTBB chorus, and a separate SATB chorus only (no solo/accomp).

CPDL would certainly benefit from more scores that were ‘right’. :grin:


There are very few scores on cpdl that I do not re-engrave for myself…

1 Like