Re: Better Midi, anyone? - dynamics
El Tue, 11 Mar 2008 12:30:07 +1100 Peter Chubb [EMAIL PROTECTED] escribió: The artemis orchestra competition has specified Lilypond as its input format since the subject of better midi on lilypond has surfaced, i'd like to take the opportunity to express a wish|request|suggestion that, being a newbie, never dared to bring up before. i think that it's far less ambitious than much of what peter is proposing. my suggestion|request is this: would it be possible that dynamic markings were mapped to velocities, and not to volume? i think having midi output at all in lilypond is a big bonus as it is, i'd be more than happy if this improvement were possible. best, lj ___ lilypond-user mailing list lilypond-user@gnu.org http://lists.gnu.org/mailman/listinfo/lilypond-user
Re: Better Midi, anyone? - dynamics
I faced the same problem four or five years ago, and I ended up writing a perl script that converted the volume changes into velocities in the midi file. If necessary, I can dig up and try to find it in my archives... Darius. luis jure wrote: El Tue, 11 Mar 2008 12:30:07 +1100 Peter Chubb [EMAIL PROTECTED] escribió: The artemis orchestra competition has specified Lilypond as its input format since the subject of better midi on lilypond has surfaced, i'd like to take the opportunity to express a wish|request|suggestion that, being a newbie, never dared to bring up before. i think that it's far less ambitious than much of what peter is proposing. my suggestion|request is this: would it be possible that dynamic markings were mapped to velocities, and not to volume? i think having midi output at all in lilypond is a big bonus as it is, i'd be more than happy if this improvement were possible. best, lj ___ lilypond-user mailing list lilypond-user@gnu.org http://lists.gnu.org/mailman/listinfo/lilypond-user ___ lilypond-user mailing list lilypond-user@gnu.org http://lists.gnu.org/mailman/listinfo/lilypond-user
Re: Better Midi, anyone?
Am 2008-03-11 um 03:10 schrieb Graham Percival: Well, I thought the original poster was offering to improve lilypond's midi export. So this _would_ work on any lilypond file... after a certain version number. :) Besides robotics I'm very much interested in enhanced MIDI output, but can't do that myself. Perhaps we can find an able programmer if we can collect some funding? I'd be in with 50-100 Euros. (That's not much, but I need these features only for my hobby projects.) As far as I understand the core devs' capacities are exceeded? Greetlings from Lake Constance --- fiëé visuëlle Henning Hraban Ramm http://www.fiee.net http://angerweit.tikon.ch/lieder/ https://www.cacert.org (I'm an assurer) ___ lilypond-user mailing list lilypond-user@gnu.org http://lists.gnu.org/mailman/listinfo/lilypond-user
Better Midi, anyone?
Hi folks, The artemis orchestra competition has specified Lilypond as its input format (see https://www.artemisia-association.org/artemis_orchestra ) for robotic instruments. We're trying to put together an entry (our robotic violinist was entered last year, with some success; (http://www.nicta.com.au/news/previous_releases3/2007_media_releases/syo_violinist_performs_with_award-winning_robofiddler ) Anyway, to the point. We want to use MIDI as an intermediate language for controlling the robot, but Lilypond's midi output ignores most of the input. For example, articulation (staccato, tenuto, accents and slurs), dynamics (I can't work out how to get a smooth crescendo on a single note), and ornaments (trills, mordents, turns etc., are not expanded). Before I start working on any of this, is anyone else doing anything in the area? Most of it may be doable by scheme scripts inserted into the source file before calling Lilypond. Harder stuff is interpreting the purely textual annotations. For example, `poco rall', `molto rit.', `a tempo', 'Tempo I', 'estinto', 'sotto voce', or (some of my favourites, from a piece by Messaien) `perdu', or `comme oiseaux'. It'd be possible to translate some of these into metronome markings or \ or similar; others, I have no idea about. Non-notated repeats are also going to be hard (`dal segno al coda', `dal segno al fine') -- 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 lilypond-user@gnu.org http://lists.gnu.org/mailman/listinfo/lilypond-user
Re: Better Midi, anyone?
On Tue, 11 Mar 2008 12:30:07 +1100
Peter Chubb [EMAIL PROTECTED] wrote:
Before I start working on any of this, is anyone else doing anything
in the area? Most of it may be doable by scheme scripts inserted into
the source file before calling Lilypond.
No, nobody is working on MIDI output. We welcome contributions to
this.
Harder stuff is interpreting the purely textual annotations. For
example, `poco rall', `molto rit.', `a tempo', 'Tempo I', 'estinto',
'sotto voce', or (some of my favourites, from a piece by Messaien)
`perdu', or `comme oiseaux'.
It'd be possible to translate some of these into metronome markings or
\ or similar; others, I have no idea about.
Non-notated repeats are also going to be hard (`dal segno al coda',
`dal segno al fine')
The above points are possibly with a macro: instead of simply
moltorit = \markup{ \italics molto rit }
create something like
moltorit = ... scheme that prints out molto rit, and tweaks
whatever options you want for your new midi code...
I'm not certain that it's worth going to quite this much trouble
-- adding articulations will likely take a few weeks. But it's
definitely *possible* to do this kind of thing with lilypond
input.
Cheers,
- Graham
___
lilypond-user mailing list
lilypond-user@gnu.org
http://lists.gnu.org/mailman/listinfo/lilypond-user
Re: Better Midi, anyone?
Graham Percival skrev:
moltorit = ... scheme that prints out molto rit, and tweaks
whatever options you want for your new midi code...
That won't work.
If I read the rules of the artemis context correctly, the solution must
work for any lilypond file.
I assume that the organizers are using some sort of simplified
lilypond-structure to represent music, so that e.g. rall. always will
be entered the same way - e.g. with -rall. and not \markup{\italics
rall.}.
Probably there is also some way of identifying different voices, etc.
But the rules do not really specify any of this, afaics.
I really do not understand why they choose full-blown .ly as the format
for this competition.
-Rune
___
lilypond-user mailing list
lilypond-user@gnu.org
http://lists.gnu.org/mailman/listinfo/lilypond-user
Re: Better Midi, anyone?
Graham == Graham Percival [EMAIL PROTECTED] writes:
Graham The above points are possibly with a macro: instead of simply
Graham moltorit = \markup{ \italics molto rit }
The point is that for the Artemis competition we have to start with
unmodified Lilypond input. So we have to recognise, say
c'^rit. or c'^rall etc., and convert to \tempo 4=60 or
whatever.
I think I can do that.
The hard dynamics one is that at present to notate a smooth
crescendo/decrescendo on a single note you have to do something like:
{ c'1 } \\
{s16\pp \ s4. \! s16 \ff \s16 \ s4. \! s16 \pp }
because there's no appropriate language construct.
That creates a new voice, and attaches the dynamics to it instead
of to the note. To produce good midi output, you need to change this
to something like,
{ c16 \pp \ ~ c16 ~ c16 ~ c16 ~ c16 ~ c16 ~ c16 \! ~ c16 \ff ~ c16 \
~ c16 ~ c16 ~ c16 ~ c16 ~ c16 ~ c16 \! ~ c16 \pp }
and even that doesn't give as smooth a transition as one would like.
Any ideas?
--
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
lilypond-user@gnu.org
http://lists.gnu.org/mailman/listinfo/lilypond-user
Re: Better Midi, anyone?
On Tue, 11 Mar 2008 03:07:17 +0100 Rune Zedeler [EMAIL PROTECTED] wrote: Graham Percival skrev: moltorit = ... scheme that prints out molto rit, and tweaks whatever options you want for your new midi code... That won't work. If I read the rules of the artemis context correctly, the solution must work for any lilypond file. Well, I thought the original poster was offering to improve lilypond's midi export. So this _would_ work on any lilypond file... after a certain version number. :) Note that I haven't read the rules, so I may be way off here. And they probably specify a version (or they should, at least). Cheers, - Graham ___ lilypond-user mailing list lilypond-user@gnu.org http://lists.gnu.org/mailman/listinfo/lilypond-user
Re: Better Midi, anyone?
On Tue, 11 Mar 2008 13:08:05 +1100
Peter Chubb [EMAIL PROTECTED] wrote:
Graham == Graham Percival [EMAIL PROTECTED] writes:
Graham The above points are possibly with a macro: instead of simply
Graham moltorit = \markup{ \italics molto rit }
The point is that for the Artemis competition we have to start with
unmodified Lilypond input. So we have to recognise, say
c'^rit. or c'^rall etc., and convert to \tempo 4=60 or
whatever.
Oh. Ooooh. Yikes. Including all English / Italian / French /
German / etc variations of slow down?!
Hmm. I was thinking of something completely different here. I
agree with Rune; I don't know why they chose lilypond for this
contest -- or at the very least, why they didn't specify a subset
of normal lilypond code.
Sorry, I can't help with this.
Cheers,
- Graham
___
lilypond-user mailing list
lilypond-user@gnu.org
http://lists.gnu.org/mailman/listinfo/lilypond-user
