Carl, I've been thinking about this, and my guess is that LP uses the bar number to calculate the beat clock. And when there are repeats, MIDI bar number != engraving bar number, so it goes wrong. This is presumably why the MIDI generator in LP doesn't handle repeats, and to make it do so would, I imagine, be a major programming task and probably not worth the effort.
I'm not sure how many LP user use the MIDI output anyway, given its restrictions. Personally, I use it for proof-reading only, so lack of repeats isn't an issue. I think it needs a mention in the documentation as you suggest. I'll try to work something up - I already have one documentation issue that I promised to help with. Best regards, Peter mailto:lilyp...@ptoye.com www.ptoye.com ------------------------- Tuesday, November 26, 2019, 7:12:47 PM, Carl Sorensen wrote: From: Peter Toye <lilyp...@ptoye.com> Reply-To: Peter Toye <lilyp...@ptoye.com> Date: Tuesday, November 26, 2019 at 11:43 AM To: Carl Sorensen <c_soren...@byu.edu>, "bug-lilypond@gnu.org" <bug-lilypond@gnu.org> Subject: Re: Midi block gives errors with bar number checks Carl, I've just found the MIDI "beat clock" - 24 per crotchet. Still no bar numbers though! Hmm, I’ve been using MIDI on a MacBook using GarageBand. Apparently Apple has added a framework called Core MIDI that adds the concepts of Bars, Beats, and SubBeats to the MIDI time stream. So I always have Bars in mm work. I suppose that one could make the midi performer ignore \barNumberCheck. The only time I can imagine that would be problematic is if one were only using LilyPond to create a midi file. And I suppose that as long as it were properly documented, users would not expect it to work. Thanks, Carl _______________________________________________ bug-lilypond mailing list bug-lilypond@gnu.org https://lists.gnu.org/mailman/listinfo/bug-lilypond
