Ah, right - the number of systems on the first page didn't change, and I stopped there instead of looking at later pages as well.
I'll experiment with nopagebreak. I already have a voice with just time signatures, skips and breaks, so that'll be a handy place to put them. Thanks, Paul On 06/09/2021 10:49:36, "Jean Abou Samra" <[email protected]> wrote: > > >Le 06/09/2021 à 11:37, Paul Hodges a écrit : >>Thanks for this explanation, which puts flesh onto the understanding I'd >>been coming to. However, I can't see what your example at the end is >>doing. Commenting out the override makes no difference to the output, >>so a little explanation might help me. > >Look at the bar numbers towards the end. You'll see that more >systems are put on each page if the \override is present. > > >>What I guess I'd be interested in at this point is a way to force >>LilyPond to put the systems closer against its own (not in this case >>better!) judgement. If not, for the time being I'll just have to reduce >>the font size slightly and accept the amount of space in the middle of >>the page. > >You can always trick the page breaking engine by adding >rows of \noPageBreak in a voice with skips and \pageBreak >somewhere... > >For example: > >\version "2.22.1" > >\paper { > #(set-paper-size "a8") >} > >\new Staff << > \relative { > c d e f > g a b c > d e f g > a b c d > e f g a > b c d e > f g a b > c d e f > e a b c > } > { \repeat unfold 6 { s1 \noPageBreak } s1 \pageBreak s1*2 } > >> > > >>As far as I can see, the stretchability parameter modifies the >>positioning of the staves on the page, but doesn't alter the decision >>about whether they can all be fitted on. > >I'm pretty sure it does (it's embarrassing that I don't even >know for sure), but this is within the limits of the space >LilyPond guesses is available before line breaking. > >Best, >Jean
