But when using it in a cross-staff
example with kneed-beams, I've noticed a problem: Some stems are affected
some not. Oddly enough, when changing the global-staff-size, I retrieve
different results.
I don't have the answer to your question, but I have an observation. The function works significantly better when you use automatic beaming. Better, but not perfectly: as you will see, most but not all stems are lengthened in the example. There is other odd behavior, too. When I change the beatStructure line (in 2.15.16) to
\set Timing.beatStructure = #'(6 6 6 5 9)
the last stem of the third group is once again unaffected.
top = \change Staff = "1"
bottom = \change Staff = "2"
music = \relative c {
\time 2/4
\override Beam #'auto-knee-gap = #0
\set Timing.baseMoment = #(ly:make-moment 1 64)
\set Timing.beatStructure = #'(6 6 6 5 8 1)
\override Beam #'stencil = #(stem-change 20)
\bottom c32 g' \top eis''
c b \bottom g,
\top e'' \bottom c,, \top g''
\bottom c,,64 \top g'' e' \bottom c,, \top c''
f,,32 f''' f,,, f'''
s64
}
\score {
\new PianoStaff <<
\new Staff = "1" {
s2
}
\new Staff = "2" {
\clef bass
\music
}
>>
}
HTH,
David