Re: Appoggiature too fast

[mskala]

I'm still hoping for a response to this, if only to confirm whether others
see the change in behaviour and that it's not something unique to my
installation.  I'm pretty sure that what 2.25.19 is doing is wrong.  The
documentation still says that an appoggiatura "takes a fixed fraction of
the main note it is attached to"; that was true in older versions (with
the "fixed fraction" being, more precisely, one half of the
following notehead duration) but seemingly no longer in 2.25.19.
   https://lilypond.org/doc/v2.25/Documentation/notation/grace-notes

Even if MIDI is not top priority for development, and it may be possible
to work around by (as usual) writing and tagging separate versions of the
music for engraving and for MIDI, the regression and the mismatch
between documentation and code behaviour, seem like problems.

I also note that that same documentation page says the MIDI duration of
"grace" notes (presumably intended to refer to \grace) is 1/4 their
notated value.  That isn't and wasn't true - it's actually 9/40 on the
default settings.  I don't know how much detail on this point needs to go
into the documentation, but it's technically not a true statement.

I think 9/40 of notated duration sounds better than 1/4 for \grace, which
is no doubt why that's the default.  The fraction being slightly less than
1/4 means the onset of the grace note is unlikely to collide with a brief
non-grace note occuring in another voice before the main note that's being
decorated.  The performance of an appoggiatura is a different matter.
That varies with historical period and personal preference, but the
earlier Lilypond default is both what I want and more likely to match
historical practice in most contexts, and it usually works out to much
longer than 9/40 of notated duration.

On Tue, 10 Sep 2024, mskala@ansuz.sooke.bc.ca wrote:

> My understanding of the intended behaviour of articulate.ly and
> \appoggiatura is that the contents of \appoggiatura, whether it's a single
> note or multiple notes, will be scaled to consume one half of the notehead
> duration of the following note, excluding dots.
>
> But since installing 2.25.19, my appoggiature are playing much faster
> than that.
>
> Example code:
>
> \include "articulate.ly"
>
> \score {
> \new Staff {
>   \articulate { \appoggiatura { e'8 g'8 } c'4 r4 r2 | }
> }
>
> \layout { }
> \midi { }
> }
>
> In version 2.25.19, the entire appoggiatura plays in a little less than
> 1/16 duration in the resulting MIDI file.  Testing with earlier versions
> is difficult because the earlier versions I have handy won't run on my
> main computer anymore, apparently due to Guile issues.  I believe, but
> cannot test on my main machine, that 2.24.1 played it as expected in
> exactly 1/8 total duration; and 2.19.83 running on another machine
> certainly plays it in 1/8 total duration.
>
> Changing the c'4 to c'2 leaves the duration of the appoggiatura in the
> output unchanged, still a little less than 1/16, and that may be a clue to
> what's going on:  I think it may be applying the 9/40 grace note
> multiplier to the 1/4 total notated duration of the appoggiatura, to get
> 9/160 which is "a little less than 1/16," and using that for the timing of
> the appoggiatura instead of the documented rule of fitting it to the
> following notehead.  It may be more or less treating \appoggiatura the
> same as \grace.
>
> Is there any simple way to restore the old behaviour?

-- 
Matthew Skala
mskala@ansuz.sooke.bc.ca                 People before tribes.
https://ansuz.sooke.bc.ca/