> On 19 Mar 2009, at 09:57, Mats Bengtsson wrote: > > > One technical problem of such a general feature, is that some > MIDI > > commands that are intended to apply to all MIDI tracks should > be > > inserted in track 0, whereas other MIDI commands (that contain > a > > channel number) should be inserted in the same track as > the > > corresponding music. Another complication (at least > conceptually) is > > the different concepts used in LilyPond and MIDI and how they > are > > mapped to each other. By default, MIDI channels correspond to > Staff > > contexts in LilyPond, but this can be changed by redefining > the > > context definitions in LilyPond. A final complication is the > lack of > > standardization of MIDI. > > > > See http://lists.gnu.org/archive/html/bug-lilypond/2003- > 10/msg00069.html > > for some related information. > > One idea is to generate some other format, perhaps Scala seq > files, > which in turn can produce MIDI files. When there is > microtonal > information present, Scala can use special algorithms to assign > MIDI > channels, and also keep track of different synth capabilities. > I'm not > sure how that format stands up with respect to LilyPond > information in > general, but such a format could be developed in a way directly > that > MIDI output cannot.
That would be very useful for other reasons too. Currently I can't tune my lilypond generated Persian midi in scala because the pitchbends generated from lilypond for the microtones are ignored when producing a scale .seq file. Kees _______________________________________________ bug-lilypond mailing list bug-lilypond@gnu.org http://lists.gnu.org/mailman/listinfo/bug-lilypond