Re: Better Midi!

[Trevor Daniels]

Hi Peter

Just tried your articulate.ly.  Looks really promising!  A couple of early 
comments.  It may be obvious, but \articulate should only be used in a 
\score block with just \midi {} - it mucks up the printed output if used 
with \layout { }.  And dynamics in the midi score block must be applied to 
all voices to be effective.

Should we consider including articulate.ly in the official releases?

Trevor

ps I like your clear well-commented code!

----- Original Message ----- 
From: "Peter Chubb" <address@hidden>
To: <address@hidden>; <address@hidden>
Cc: "Glenn Downey" <address@hidden>; "Diana Nguyen" 
<address@hidden>; <address@hidden>; "Dorothy Kennedy" 
<address@hidden>
Sent: Wednesday, August 06, 2008 4:38 AM
Subject: Better Midi!


> Hi Folks,
> I now have permission to release the articulation code
> to improve MIDI output for lilypond.  I've attached it as a
> tar.bz2 file here.  The code is copyright 2008 NICTA (my employer),
> but released under GPL version 2.
>
> The simple way to use it is to use the lilywrap script in the tarball:
>  $ lilywrap inputfile.ly
>  $ timidity inputfile.midi
>
> Long version of how to use it:
>
>  \include "articulate.ly"
>
>  Insert \unfoldRepeats \articulate into the appropriate places.
>
> Short Example:
>
> \include "articulate.ly"
> \score {
> \unfoldRepeats \articulate <<
> \context Staff {
> \set Staff.midiInstrument="clarinet"
> \relative c' {
> \time 4/4 \tempo 4=100
> c4-. c4( d e |
> f\trill) g\turn a b |
> c1-- 
> }
> }
> >>
> \midi {}
> }
>
> What it does:
> * Any note not under a slur or phrasing mark, and not marked with an
>   explicit articulation, is shortened by ac:normalFactor (default
>   7/8)
> * Any note marked staccato is shortened by ac:staccatoFactor (default
>   1/2)
> * Any note marked tenuto gets its full value.
> * Appogiaturas are made to take half the value of the note following,
>   without taking dots into account (so in \appoggiatura c8 d2. the c
>   will take the time of a crotchet)
> * Trills and turns are expanded.  The algorithm tries to choose notes
>   within the time of the current tempo that lead to each twiddle
>   being around 1/8 seconds; this can be adjusted with the
>   ac:maxTwiddleTime variable.
> * rall, poco rall and a tempo are observed.  It'd be fairly trivial
>   to make accel. and stringendo work too.
>
> There's a TODO list a mile long; unfortunately I'm no longer being
> paid to work on this, but I'm happy to coordinate the attempts of
> others to work on it.  Also, my scheme coding is appallingly bad (this
> is the first serious work I've done in scheme) so there're probably
> major improvements that can be made.


--------------------------------------------------------------------------------


> --
> Dr Peter Chubb  http://www.gelato.unsw.edu.au  peterc AT 
> gelato.unsw.edu.au
> http://www.ertos.nicta.com.au           ERTOS within National ICT 
> Australia


--------------------------------------------------------------------------------


> _______________________________________________
> lilypond-user mailing list
> address@hidden
> http://lists.gnu.org/mailman/listinfo/lilypond-user