RE: Percussion!

[email protected] ("Lee Goddard") Sat, 19 May 2001 10:56:46 +0100
Newsgroups perl.midi
Message-ID <[email protected]>
> On Sat, 19 May 2001, Lee Goddard wrote:
> > Yes, but let's at least make sure it's musical in some sense.
> 
> imho music is just "organised sound."  Perl is good for that i think.

You must be American, sir?
 
> > Reich would be relatively easy, don't you think?
> 
> why not try to find new ways of organising sound rather than reducing our
> use of computers to mimicking others?

Please see above.
 
> > thanks again for the libraries
> 
> Computers're bringing about a situation that's like the invention of
> harmony. Sub-routines are like chords.  No one would think of keeping a
> chord to himself. You'd give it to anyone who wanted it. You'd welcome
> alterations of it. Sub-routines are altered by a single punch. We're
> getting music made by man himself: not just one man.
> John Cage, 1969

If you punched my sub-routine, sir, I'd punch you too.

In my experience, many many musicians keep chords too themselves.

Cage is obviously out of date.
 
Seriously though (and it hurts to be serious): perhaps I was less
than lucid in expressing my point, why seem to be supporting.

Traditional AI approaches to making 'original' 'music' have made what 
the vast majority of listeners refer to as 'sound' or 'noise' (the
earliest I know of being ILLIAC, by Hiller and Isaacson).

There are many so-called successes in mimicry - the best seeming to
be folk-song writing programs - that generally use a logic map to
allow certain progressions (like, from II we can go to IV, V only).

Improvisation around 'jazz' rifts is also apparently rather easy:
Fry's Flavor's Band, for example, or the chap at Sussex cf. Pine.

My interest is less practical, more philosophical, and so less
profitable (in the short term: in the long term, of course, I'll
be Andrew CCoyd Wedder).  I've only ever heard of a chap called
Moore suggesting an investigation of the mathematical relationship
between waveforms in the sense of tonal harmonies and rhythms.

There are plenty of neural networks that discover mathematical 
relationships, and there are plenty of mathematical relationships
that the vast majority of people find pleasing, as if they 
represented some Platonic Ideal manifest.

From these comes a phrase that can be processed by a genetic 
algorithm to produce a variation, the whole being refed into
the NN to create a second part, and so on.

Maybe.

Send me money and I'll tell you more.

Lee
(Pronounced 'pretending')