Greetings everybody, hi Nicolas,
I'm not sure anybody else has encountered this before (couldn't find
any such reports in the tracker or in bug- archives). Have a look at
the following example, where all four fragments should look the same
(hint: they do not).
In the first two, Lily behaves as expected. In the third and fourth,
however, the Scheme function used to generate the polyphony confuses
the chord-repetition mechanism. See attached output.
\version "2.14.0" %% or so :-)
polyTestOne =
#(define-music-function (parser location lower upper)
(ly:music? ly:music?)
#{<< { \voiceTwo $lower } \\
{ \voiceOne $upper }>> #})
polyTestTwo =
#(define-music-function (parser location lower upper)
(ly:music? ly:music?)
#{<< { \voiceOne $upper } \\
{ \voiceTwo $lower }>> #})
\relative c' {
<d e>2 q
<<
{\voiceOne<d e>4<d e> q q} \\
{\voiceTwo g,2 r}
>>
}
\relative c' {
<d e>2 q
<<
{\voiceTwo g, r} \\
{\voiceOne<d' e>4<d e> q q}
>>
}
\relative c' {
<d e>2 q
\polyTestOne {g, r} {<d' e>4<d e> q q}
}
\relative c' {
<d e>2 q
\polyTestTwo {g, r} {<d' e>4<d e> q q}
}
%%%%
Cheers,
Valentin.