On Thu, 03 Jun 2004 15:42:33 +0100, alex wrote
> On Tue, 2004-06-01 at 18:21, Dave Griffiths wrote:
> > There is a simple score language the lsystem produces where:
> > o = note on
> > + = up one note (the exact frequency change is configured elsewhere)
> > - = down one note
> > ! = reset to root note
> > . = rest one beat
>
> Nice!
>
> My most successful attempt at making a tune is similar to this. I just
> use a two dimensional structure of +'s and -'s, and step through
> structure recursively.
Sounds cool, but I don't understand fully... Most of my attempts at creating
melody (which work) have just consisted of sampling low frequency sine waves
for pitch and note on/off information and tweaking them to cycle patterns, but
this is getting off topic...
> I'm not sure what an lsystem is, but I think my naive approach is similar
but a lot simpler.
An lsystem is really just a set of search replace rules you iteratively apply
to a string. Most of the work is in how you interpret the symbols. (good
description here:
http://en.wikipedia.org/wiki/Lindenmayer_system)
Branching is also important, but in this system it doesn't make use of that
for the melody.
> The first bit of this recording from pescara is using live coding:
> http://nosignal.slab.org/placard/mp3s/yaxu.mp3
Cool - I think I like the livecoded bit the best :) I meant to post some
lsystem music earlier but forgot:
http://www.archive.org/download/a-long-walk-take1/a-long-walk-take1.ogg
http://www.archive.org/download/a-long-walk-take2/a-long-walk-take2.ogg
Not really strictly livecoded as I got the lsystems set up first and tweaked
them, I'll try recording some written from scratch soon...
cheers,
dave
Received on Thu Jun 03 2004 - 17:00:10 BST