Re: lilypond: grace notes and MIDI

[Heikki Johannes Junes]

On 1 Sep 2002, Thomas Bushnell, BSG wrote:

> A while ago you posted a bug report in which grace notes cause the
> different staves to become unsynchronized on MIDI output (but just
> fine on printed output).
>
> I am encountering the same problem; it seems to work fine if there is
> only one grace note, but if there are multiple ones, then the
> following note is incorrectly lengthened by the length of the grace
> sequence less one note.
>
> Have you found any work around or other fix for the problem?

No, I only reported exactly about the same error. The paper output was
fine, which is the most important. In order to proofcheck the rest of the
piece the graces had to be commented out, which is a pity if you have
groups of graces in several places. Currently, if groups of graces are
used throughout the piece then midi becomes almost unusable.

IMHO, it would be the most convenient if one could show exactly how much
earlier graces should play. For example,

   e2 | \grace{ f8 } e2
   e2 | \grace{ f16 g16 } e2

would start play exactly '8'-duration earlier in midi, but

   e2 | \grace{ f8 }*1/2 e2        % a proposed syntax
   e2 | \grace{ f16 g16 }*1/2 e2   % a proposed syntax

only '16'-duration earlier and would be played with double speed. With the
proposed '\grace{...}*1/2'-syntax above one could vary the speed of
graces. There is a lot of music which uses graces.

      Heikki Junes