Re: MIDI dynamics

[Hilary Snaden]

On 04/02/18 05:19, Abraham Lee wrote:
> Hi, Hilary!
>
> On Sat, Feb 3, 2018 at 9:42 PM Hilary Snaden <address@hidden> wrote:
>
> > I have been trying to get LilyPond to generate MIDI versions of some
> > piano music. The music script renders nicely, and there are no errors or
> > warnings, but the MIDI file plays with no dynamics at all. What am I
> > doing wrong? (LilyPond 2.18.2.)
> >
> > \score {
> >     \unfoldRepeats
> >     \new Staff
> >     <<
> >       { \tempo 4=86 }
> >       \new Voice = "v1" {
> >         \set Staff.midiInstrument = "Acoustic grand"
> >         \set Staff.midiMinimumVolume = #0.2
> >         \set Staff.midiMaximumVolume = #0.8
> >         \upper
> >       }
> >       \new Voice = "v2" {
> >         \set Staff.midiInstrument = "Acoustic grand"
> >         \set Staff.midiMinimumVolume = #0.2
> >         \set Staff.midiMaximumVolume = #0.8
> >         \lower
> >       }
> >       \new Dynamics \pianodynamics
> >     >>
> >     \midi { }
> > }
>
>
> In order for dynamics to affect the notes, they need to be directly
> attached to them. Since this score is solely for playback, instead of
> creating a separate dynamics context (which is really more helpful for the
> engraving part), make \pianodynamics simultaneous with the variables \upper
> and \lower *within the Voice contexts*. So, something like this:
>
> \new Voice { ... } <<
>    \upper
>    \pianodynamics
> > >
>
> and remove the Dynamics context completely since it doesn’t have a direct
> affect on playback.
>
> Hope that helps,
> Abraham

It does help, thanks. I found that I also had to move the midi volume 
adjustments into the staff context, so the midi block now looks like this:

\score {
  \unfoldRepeats
  \new Staff {
    \set Staff.midiInstrument = "Acoustic grand"
    \set Staff.midiMinimumVolume = #0.05
    \set Staff.midiMaximumVolume = #0.85
    <<
      { \upper }
      { \lower }
      { \pianodynamics }
      { \tempi }
    >>
  }
  \midi { }
}

For best results I'll need to add explicit levels at the end of every 
cresc and dim in the variable containing the dynamics, but this is still 
a significant improvement over no dynamics at all. Cheers!

--
Hilary