Re: MIDI staffs or voices?

[Paco Vila]

El Fri, 19 de Oct de 2007, a las 09:43:57PM -0700, Ralph Little dijo:

> \version "2.11.32"
>
> \midi {
>  \context {
>    \Staff
>    \remove "Staff_performer"
>  }
>  \context {
>    \Voice
>     \consists "Staff_performer"
>    }
> }
>
> \score {
>  {
>      <<{c' d' e' f'} \\ {c' b a g} >>
>  }
>  \midi{
>  }   }


There is another limitation of MIDI, only smart enough MIDI performers 
are able to play the following correctly (substitute it into your 
previous example):

{
     <<{d'4 c'2 b4 c'4 r2.} \\ {b4 c' d'2 c'1} >>
 }

If all voices from a staff go to a single MIDI channel, then a 
note-off cancels the sound of this note in this channel. This is a very 
good reason to assign channels to voices, not staves.

Also note that MIDI tracks can mix notes tagged with different 
channels. This way, a staff could translate to a single track with 
notes on separated channels into the same track. 

Assigning a MIDI output channel to a track is an option in some 
sequencers, but you can always leave the channel of each note untouched 
if you do not assign channels to tracks.

Could anybody test how this example sounds? c'2 in first measure should 
sound as a half note together with part of the d'2 in the other voice; 
also, c'1 in the second measure should last as a whole note, not a 
quarter note.

--
F. Vila