INTERVIEWS, REVIEWS & RELATED ARTICLES
"Craft is what enables you to be successful
when you're not inspired." - Brian Eno
Q JULY 1996 - by Robert Mills
MY EVER CHANGING MOOGS
Fed up with your CDs just being, like, the same every time you play them? Bo-o-ring! Well, Professor Brian Eno has patented Generative Music, that, quite simply, forever changes. He demonstrates the eternal remix to a boggling Robert Mills.
Flying toasters? Worms? Tropical fish? What on earth have this little lot got to do with Brian Eno and a new and intriguing musical innovation? Screen-savers is the short answer. Oft trivial, but nevertheless visually compelling, these tiny, self-activating computer programs not only prevent monitor-wrecking image burn-in over extended coffee breaks, but have also, in some small way at least, inspired the latest in a long line of Eno's conceptual and technological experiments.
The studio mantelpiece chez Prof begs space for another Q Inspiration Award as the archbishop of ambient turns evangelist, presenting a software-led system that builds upon ideas he dabbled in during the mid-'70s, and offers music technology what promises to be another exciting new avenue. Eno's attraction to what he's calling Generative Music is rooted in a long history of experiments dating back to his 1975 release, Discreet Music.
In those days, Eno's simultaneous spooling of pre-recorded tape loops of various lengths through separate amp and speaker combos conjured seemingly never-ending permutations of music, thanks to the gradual breakdown of the tapes' synchronicity. Eno cites an attempt to combine five cassettes, with loop lengths of 23, 25.5, 30.2, 19.7 and 21.3 minutes; the music would theoretically only ever drift back into sync after fourteen and a half years had elapsed, thus creating, at any intervening point, a truly unique musical experience.
An ensuing collection of homespun recordings known as the Ambient series (Ambient 1: Music For Airports and its successors) were all shaped with this technique, but, as far as Eno is concerned, were fundamentally flawed by the fact that the listener heard the same forty-minute extract each time.
Now, almost twenty years up the technological ladder, Generative Music 1 brings with it a completely original way of listening to music. What you get is the inherent spirit of a live performance but with the praiseworthy convenience of a hi-fi. And because the whole thing is run from a standard 3.5 inch floppy disk, there's no expensive hardware to shell out for.
Centred around the relationship between a pioneering form of musical-authoring software - of which more later - and the personal computer's sound card, it's now possible for the computer processor itself to compose streams of perpetual unique melodious merriment.
Poised before what appears to be a newly installed PC at his unexpectedly ice-cool West London premises, Eno is visibly enchanted with the system's undeniable self-propelling focus and by the kind of music it allows him to engineer. This is the first thing of its kind available, he gushes, I love the economy of it; I can get the biggest effect for the smallest amount of input. It just carries on with a life of its own. I've always liked the idea of setting up a number of processes and letting them work themselves out. Clearly, the consumer's experience is dependent on the way each strand is set up, and this is where the brain-power of the computer chips in.
To ink in a little background, in February 1995, Sseyo Ltd (a software development company) sent Eno a selection of music created with their Koan Pro software. The results were very much in the vein of Eno's ambient work. It was a humbling experience, he concedes.
But the reason I'm so evangelistic about this process is that it's not just interesting in musical terms; it's a way of making computers do what computers do so well. Eno is critical of the way much computer technology has failed to fulfil its potential. I despise computers in many ways. I think they're hopelessly under-evolved and overrated, he sighs. People jumped into CD-ROM thinking it would replace existing systems; this was supposed to be the paperless office. The reality is that the old technology carries on, and in some cases gets better. When CD-ROM first appeared, everyone was very hyped up about it and thought it was the future of everything. I looked at the available product and thought they were anything but the future, and possibly the most serious technological error.
These are near-heretical statements, especially given the current grip of PC fever, and although Generative systems triumph in their ability to handle a spectrum of probabilistic conditions, the fixed nature of the ROM couldn't be less helpful.
The thing I dislike so much about CD-ROMs is that they use the brute force of the computer - the ability to move huge blocks of data around, such as displaying pre-defined pictures and information. With this, instead of using the computer as a means of 'showing', the idea of using computers as a tool for 'growing' things takes over, like some of the advanced forms of animation or the study of cellular automata. That's what scientists are using them for. Look at what scientists are doing and the potential of this tool has simply not been touched. It's a convincing justification, applying to both composer and end user.
For Eno, the beauty of the KOAN system is that the architecture of the music can be layered empirically, which means the outcome is heard as it's being written. The authoring software allows him to allocate any of around a hundred and fifty different conditions, each operating within its own probability range, to any given voice or instrument. Clusters of instructions are then directed to the PC's sound card to produce a signal for replay. Subtle differences between note timbre, shading, rests between notes - not to mention scales and their associated musical harmonies - can all be apportioned a distinct range of possibilities. The creative permutations are endless, and for Eno, at least, highly addictive.
Sometimes I have ideas in the night. I wonder what would happen with these two rules, or that particular arrangement. One of my favourite pieces was made in about six minutes, but I must have listened to it for hundreds of hours. These musical seeds can be rigid enough to give each individual composition its own identity or designated with a broad probability quotient, whereby the variation between each replay session will be magnified. There's one little sequence that I've only ever heard twice. It's a note that can only play when every other condition is right. It's like hiding a little jewel - when it comes along it's a real joy.
Of course, each set of conditions can interact with other sets to create hybrid situations, meaning, reckons Eno, that truly credible computer interaction is possible for the very first time. For interaction to be meaningful you have to have very subtle controls. If you tell someone that interaction is just switching something on or off or making a choice between three things, that's the most pathetic form of interaction imaginable! spits Eno likening his new toy to a kind of space-age wind chime, with dozens of contrasting elements working within fixed and flexible constraints, with the wind acting as the constant, which also has its own natural variants.
It's certainly not random music, he assures. I would say it's more organic, as opposed to mechanistic. He illustrates his point with an impromptu demonstration. You need to leave the thing on for a while to get the measure of it, although there is always a high enough surprise level to make it worth keeping on all day. It's constructed in such a way that there's a good chance you're not going to hear the same thing more than once.
There's one piece on the disk (Seed Reflector, as it turns out) with very little melody but a lot going on texturally that gets refreshed every two minutes or so.
More pertinent, perhaps, is the fact that the system is not limited to producing the kind of material that Brian Eno is renowned for. There's a whole area of things that I didn't do. You can make convincing techno music, which is just not the kind of thing I feel like doing. Other musical genres which operate within well defined rules are also fairly easy to simulate, as Brian himself found out.
I've composed some pretty convincing Schoenburg on it, he confesses. I played it to David Bowie and he was so thrilled by it, he suggested we could invent an early twentieth-century composer, who's just been discovered, and start putting out records by him.
Although the biggest noises in the PC arena are currently being made with even bigger and more powerful machines, Eno is satisfied with the potential of this relatively simple system. This is a little signal as to what could be done with it and what I keep saying to people is, When you listen to this piece of music, whether for five hours or five minutes, it's just 17K of memory. It's astonishingly tiny. When you think a CD is six hundred megabytes for just one hour of music, the comparison is infinitely in favour of this.
At the moment, of course, the system's appeal is narrowed to PC users, and the software itself only permits the compositions to be played, rather than played with, but Eno is already thinking about next generation of possible music systems which could broaden its exposure.
I can imagine the future: your hi-fi will be a mixture of a player, but also a grower. You could release a CD with both fixed and free material so that the machine both makes music and replays it at the same time. What you'd have is an eternal remix. Now, I could really see that catching on...
A tad confused? Don't be. For the uninitiated, a sound card is a special kind of circuit board which functions like a mini-synthesizer, capable of turning out a range of voices and sound effects. Located in one of the PC's expansion slots, it works behind the scenes turning software commands into the very real sounds you actually hear. It makes sure the correct sounds are played at the right time, and is thus essential in the modem scheme of multimedia things. Imagine playing your favourite shoot-'em-up game without its backdrop of raging metal fretwork and full-width explosions; the audio soundtracks of CD-ROMs too would be silent without one and consequently all new PCs designated for multimedia applications (i.e. nigh on all of them) come with said card fitted as standard.
The Brian Eno Generative Music 1 floppy arrives with two versions of the Koan player software; both Windows 3.1 and Windows 95. Either will work fine and dandily under the processing muscle of an IBM compatible 486 chip (or higher), with at least eight megabytes of RAM. What's more crucial to observe is the precise type of sound card that's installed. This product is designed to work exclusively with Creative Labs AWE32 or SB32 cards or the TDK MusicCard. Luckily, that covers most recent models, including the popular SoundBlaster family.
Alas, no Macintosh edition is available: Macs are so damn clever they don't need a special sound card to play music and effects. Unfortunately, Koan relies on the sound card, so Mac addicts will have to buy from Eno's back catalogue and set their Apple CD Audio Players to infinite repeat.