From: "Hugo Sweet" <[EMAIL PROTECTED]> To: "313 (E-mail)" <313@hyperreal.org> Sent: Tuesday, November 13, 2001 10:31 PM Subject: [313] Art and Technology again
> In my opinion, and based purely on the above, software such as EMI could not > produce great techno because of the priority given to the characteristics of > the sounds used (especially where effects and "found" samples are > concerned), rather than the rules of musical theory behind their > arrangement. It could be argued that techno is the distillation of music > theory, a music that goes beyond the limitations of acoustic musical > technology. Once music theory has been distilled to variations upon the > theme of the heartbeat, the sound palette can be set free. Electronic music > is therefore the inverse of classical music, and other music based on a > restricted palette of acoustic sounds. Where in classical music it was the > arrangement that defined the style of a composer, with techno it is the > sound palette itself. Only in techno does it make sense for a composer to > say that using a preset synth sound is either succumbing to cliche or is > creative cowardice. If a computer can learn how to write music, I fail to see why it couldn't also learn how to produce music. The characteristics of sound could be analysed as easilly as the notes played. Samplers and software that we use today already do this very well. To be fair, you could also feed it the arrangement and sequencing of the song, and "teach" the computer to look for correlations between the originally sequenced track and the final audio (if all the original knob/fader movements were recorded into MIDI and the effects were DSP). You could teach it about the evolution of synthesis and production, and all the latest gear. Given that you'd be working with audio samples rather than MIDI, the time it would take to learn would be that much greater, and the variables are larger, but given time, I'm sure it could be done to the same degree that "the classics" can be learned. And to the same degree, one computer might limit itself to familiar tools, while another might tend to explore new options. One might generate chaotic noise, another might generate a symphony. If we put 1000 computers in a room and played 1/100 techno songs, and 50/100 trance songs, most of them would probably make trance. Alright, maybe that was silly, but you catch my drift. Musical taste is learned by humans, starting with what we're spoonfed. Who knows why we chose what we chose beyond that? Arguments to the "objective superiority" of style X hold no water. So if it's not objective, what is it? where does that root of choice come from? What in humans distinguishes our faculty of taste from that of a computers? What is taste? What is it that makes "new" music new? What is random? What is creativity? What is spontanaeity? These things have been debated ad nauseum to no conclusion. Give the computer some Hendrix, King Tubby and Kratwerk and see where it winds up 20 years down the road. Moreover, give a few hundred thousand computers 20 years in their own vaccuum to share musical ideas from different musical influences, merge that with programmers feeding them new instruments, and see where they wind up in twenty years. I think Mike's comments the other day were great. Take this scenario and *combine* it with the shared experience of music humans own today, and together there could be great developments. I'm reading Bergson right now, and this all fits far too well. With AI innovations like these, I see programmers leaping from assigning narrow tasks (like an amoeba) to giving computers the faculties to *link* distinct work and generate a response, given a broad pallette. Humans have a very developed network of learned associations that form complex perceptions of the world, compared to microorganisms that sense little, and respond within a narrow framework. As the task of organizing a complex array of sensation is combined with a larger memory, the numbers of options for any given action expand. A child may form a concrete dislike from one bad hot dog, and as they grow older, they may experience more hot dogs in different ways and based on those experiences find new arrangements of condiments that will become identifiable with their taste. With both computers like EMI, and adults eating hot dogs, we are unable to determine what choice they may make, and with both complex thought systems, we are unable to gleen the root of that spontanaeity. It wouldn't be spontaneous if it had a root, right? I'd say we're already as unable to determine the roots of choice in complex computer programs as we are with humans. Why did EMI chose the notes it chose? If we keep broadening its pallette of choice, I can see vast potential for creativity. Of course, this is all a huge oversimplification, b/c computers are not presently predisposed to one action or another through a history of experience, they aren't fed experiences outside of the realm of a few musical passages from famous composers and they don't play them top-40 radio either. But since this is just the beginning, I think some argumentative slack can be granted. Also, this all begs a discussion about spirit and soul and what-not, but you can tell where I stand on that fence. Tristan ---------- http://ampcast.com/phonopsia <- Music http://phonopsia.tripod.com <- Mixes, pics, thought, travelogue & info http://www.metatrackstudios.com [EMAIL PROTECTED] <- email <FrogboyMCI> <- AOL Instant Messenger --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]