Re: Map voices to channels in MIDI output

[Jan Nieuwenhuizen]

Reinhold Kainhofer schreef op ma 14-03-2011 om 14:42 [+0100]:

> Yes, but my main concern is not to use midi2ly on files generated by lilypond 
> in the first place.

Fair enough.  My concern is mainly the MIDI->ly traject, in combination
with adding a combined test suite.  Until now, MIDI has been completely
untested, I think.

> My concern is to use the midi file to listen to the music 
> and eventually generate nicely sounding output. Having one instrument spread 
> out to 5 midi channels (1 for the non-partcombined sections and 4 for the 
> part-combined sections) does not help, when you want to adjust some aspects 
> of 
> the playbeck, since you then have to apply the identical settings 5 times...

Each voice goes to its own track, however, they all go to the same
channel.

> > Does midiChannelMapping == #'staff still produce the old output?
> 
> No, I couldn't get it to work.

Ok, thanks.  There was a bug, 'staff-mode still created
multiple tracks.  Fixed now.

> See the attached file.

Hmm, this looks as a real weird piece of music, having partcombine
kick-in mid-staff?  Why not something like

\score {
  <<
    \new Staff=StaffI <<
      \set Score.midiChannelMapping = #'staff
      \new Voice=one \relative c'' {
        \partcombine {
          \key g \major
          \time 6/8
          d4 b8 c4 a8 | d4 b8 g4 r8
          \time 4/4
          c4 d e f | g a b c
        }
        { R1*2*6/8 | c4 d c d | e8 f g a r2 }
      }
      >>
    \new Lyrics=LyricsI \lyricsto one {
      Girls and boys come | out to play,
    }
    \new Staff=StaffII \new Voice=BBBB \relative c'' {
      \key g \major
      \time 6/8
      d4 b8 c4 a8 | d4 b8 g4
      r8 R1*2
    }
  >>
  \layout {}
  \midi {}
}


Incidentally, how did you produce the .png?  It looks useful, apart from
the fact that according to midi2ly, there is really only one channel
being used -- I cannot confirm this image.

> I'm currently not able to reproduce the lyrics problem with a simple file 

Good :-)

> But the part-combined voices are all placed 
> at the very bottom. See attached sample case.

Should be fixed for 'staff mode and I don't think that should hurt in
'instrument mode.  Also, I don't think this example is very realistic.

Greetings,
Jan.

-- 
Jan Nieuwenhuizen <address@hidden> | GNU LilyPond http://lilypond.org
Freelance IT http://JoyofSource.com | Avatar®  http://AvatarAcademy.nl