A few helpful hints after an accidental tweak.
First off, I noticed something interesting: the nasty chords ALWAYS appeared the first note following a completed \repeat volta n. Each chord occurs only once, even if it itself is in a \repeat block.
Because lily doesn't unfold repeats automatically, I've been using two score blocks as recommended in the manual: one for layout, one for MIDI. The MIDI score usually looks like this:
\score {
\simultaneous {
\new StaffGroup <<
\unfoldrepeats {
\new Staff <<
\set Staff.midiInstrument = #"flute"
\new Voice { \Fluteone }
>> }
\unfoldrepeats {
\new Staff <<
\set Staff.midiInstrument = #"oboe"
\new Voice { \Oboeone }
>> }%%% etc.
There may be a more efficient way of doing this, but the only way I managed to unfold the piece completely was by applying it to each instrument individually.
Anyway, because this project has been collecting dust since 2.3, I put a little effort into making it into a \book. For simplicity, I was required to remove the MIDI score block and replace it with \midi { } in each score. After compilation, I checked the new MIDI...the chords were gone! Alas, so were the unfolded repeats.
HTH, Will
On Apr 23, 2005, at 2.59 PM, Erik Sandberg wrote:
On Saturday 23 April 2005 17.39, Will Oram wrote:I understand your concerns. If I had any more specific information, I would have gladly posted it...this is by far not the first time I've described a bug on bug-lilypond. I've tooled around with lily for...wow, 1.5 years already. Anyway, I have done my absolute best to identify the root of these problem chords (and it ain't an inverted V, haw haw), but it seems to be a truly random problem.
Seems like a strange bug. If you want to isolate it, here are a few ideas you
can experiment with:
- Change the pitches globally (with search/replace), to just include a few
ones (say c, e and g). Does this change the problem? If not, do the strange
chords change pitch when the other notes do? In that case, which notes
control the pitches of the strange chords?
- How does the problem change if you remove one bar at the end or beginning of
the piece, or if you remove one staff? How much can you remove before the
problem disappears?
- You can use midi2ly on lily's output to view the midi files' content
- When you have done the above experimenting, I'm sure you have reduced the
problem to something whose .ly source you can share.
- You could try 2.4's midi output also, 2.5 is considered unstable.
good luck! Erik
I don't dare succeed
if change always comes first
----------------------------------------------------------
_ | WILL ORAM
ASCII ribbon campaign ( ) | spamguy (at) foxchange . com
- against HTML email X | wro1 (at) cwru . edu
& vCards / \ |
| AIM spamguy21
----------------------------------------------------------
(ABOVE: Magnetic poetry #5)_______________________________________________ bug-lilypond mailing list bug-lilypond@gnu.org http://lists.gnu.org/mailman/listinfo/bug-lilypond
