Hi Robin, thank you very much for your input!
On Tue, Jan 6, 2026 at 12:42 AM Robin Bannister <[email protected]> wrote: > Heiko Schill wrote: > >> > >> Is there a way to get the desired output somehow? > >> > > Your code starts by defining the music for each voice. > Fair enough. Actually, this is mostly the code of Eirik or yourself borrowed from the recent thread you cited below :) I found that thread after preparing my own MWE which looked mostly similar but was not as minimalistic as this one... > But these definitions seem to expect that Cello will be second and Voice > will be third, the situation set up in the first score. > Yes, this is the order of appearance. In the first example, though, the order of the staves is not what I would like to see. The second score defines a different order for the staves, but uses the > voice definitions unchanged, invalidating their expectations. > The third score has the same muddle, labeling the d1 sequence 'Voice'. Sure, this is because I wanted to keep the example minimal. I set up the labels correctly in the "real" score. Your tip using \markuplist { "1" "2" "3" "4" "5" "6" "7" "8" } from https://lists.gnu.org/archive/html/lilypond-user/2025-12/msg00126.html is really helpful in more complex situations. My actual problem is something completely different - sorry for not having explained that well enough before: In the second example, the engraving is bad when the first "new" voice, the cello, starts in m.11 (see attached picture). The clefs is m. 11 are shifted correctly to the right and the notes follow suit, but the staves themselves start on the far left where any non-indented staff would start. Additionally, no label (not even in the wrong position) is produced. This seems to happen, when the order of appearance does not match the order of the staves. In the first example, the first voice to start (piano) is engraved as the first staff, the cello as the second instrument to appear is on the second staff and so on. This works as expected. If I only change the order of staves to the desired layout (1st staff: voice, appears last, 2nd staff: cello, appears second, 3rd staff: piano, plays from m.1 through to the end), I obtain the output for the second example. The third one is identical to the second, but just omits the \RemoveAllEmptyStave. This also works as expected (correct indentation of the staffs, all labels engraved as intended, although in this example mixed up because the \markuplist is not adjusted), but produces a massive amount of full measure rests: The whole piece consists of 100 measures, the cello plays throughout m.17-72 and the vocals only start in m.73 through to the end. You can actually easily create an example with four (or maybe even more, haven't really tried) staves per system in the same fashion and you will see the same situation as in m.11 for all but the last occurrence of \pseudoIndent in the lilypond code. So the question actually is: Why does \RemoveAllEmptyStave screw up the indentation of the staves in example 2, but not in example 1? Is there a way to get the cello label in front of m.11 in this example and the staves starting in the same horizontal position as the ones in m.21 even if the order of the staves is like those in the second example? Best regards, Heiko
