Dear bug squad, I am rephrasing a report that got lost in a long-ish thread <http://lists.gnu.org/archive/html/bug-lilypond/2010-12/msg00403.html> discussing whether it was really a bug. The thread eventually made clear: it is.
% When there is a clef change within a measure, we want cancellation
accidentals to
% be printed, if they would be printed for the same pitches without the clef
change.
\version "2.12.3" % 2.13.43 behaves the same
\relative c' {
\key d \major
\clef bass
c4
\clef treble
c4
\clef bass
cis2 } % the sharp symbol is missing from the output
% Workarounds: #(set-accidental-style 'piano)
% or inspect the output near clef changes
% and force accidentals, cis!2, when required
The regression test accidental-clef-change.ly demonstrates that accidentals
moving
away from the key signature are repeated after a mid-measure clef change. This
is
helpful, but might cause the problem above as a side-effect.
I would consider adding the extra accidentals in accidental-clef-change.ly to be
not as important as printing the within-measure cancellation accidentals.
The best behavior would be to have both: print accidental symbols if they would
print
in the chosen accidental-style either in absence of the clef change, or with the
clef change treated as a reset to the key signature.<<attachment: missingAccidental.png>>
_______________________________________________ bug-lilypond mailing list [email protected] http://lists.gnu.org/mailman/listinfo/bug-lilypond
