Re: Was good, now assess


Subject: Re: Was good, now assess
From: Kevin Austin (kevin.austin@videotron.ca)
Date: Sun Dec 16 2007 - 18:51:37 EST


Well, yes.

At 5:02 PM -0500 12/16/07, Eliot Handelman wrote:
>Well, at our university we make the following spiel. We are at the
>beginning of a brand new epoch in music. What will come out of it is
>extraordinary new music & extraordinary new musical experiences.

Or perhaps extraordinary new perceptions of old musical experiences.
Such as being able to experience the Webernian canonic retrograde
inversions in realtime.

>In order to cross the threshold into the new, however, we need to be
>able to give computers ears & musical minds.

I remember when people said "Hey! 4k (bits) of memory, let's do some
AI." The first generations of 8 co-processor desktops are starting to
appear. I think that we may need to get into a couple of generations
of 64 co-processors to start to do any of the 'real' things Eliot
proposes. With 8 parallel processors, it's still not possible to
model real reverb ... still sounds like a machine.

>We cannot do serious musical computing without this.

(1) The machines don't exist. (2) The models for what to do with
these machines don't exist.

>However, this is one of the most complex and abstract problems that there is.

It's not going to be easy, but it may be simple.

>Thus we must prepare.

Agreed.

>We are all obliged to sit through Eliot's semester-long lectures on
>Happy Birthday,

Education does not occur through being lectured. Ask any teenager.
Does Eliot have insight into the melody Happy Birthday?

>courses in the analysis of folksong,

Would this include the differences between North Indian and Afghani
music? The impact of prosody and global migration?

>computational models of Beethoven, a course on bad music (one of our hits).

Everyone hits on bad music.

>-- eliot

Good luck. Sounds well worth what it costs.

Best

Kevin



This archive was generated by hypermail 2b27 : Sat Dec 22 2007 - 01:46:31 EST