In Part One, Nigel Jopson talks computer music to Peter Zinovieff, the creator of Electronic Music Studios.

The largest and most sophisticated computer-controlled electronic music studio in Britain sits in a wooden shed in a rambling Oxfordshire garden, surrounded by itinerant sheep and chickens. The owner is Peter Zinovieff, founder of EMS, the British synthesiser manufacturers, and he greets strolling journalists like myself with a generous spread of home-cooked food for lunch and a well-aimed barrage of rotten eggs as I approach the music studio - which costs somewhere in the region of £20 000 (and God knows how much in man hours of work and worry) a year to maintain and develop, produces music only 10% of the time (Peter's estimate), and which he gave to the music faculty of Oxford University earlier this year.

If you examine the photos that accompany this article, or if you have already seen EMS's publicity pictures of the studio at its previous location in Putney, then you will realise how truly incongruous its present setting is; if you read this article, which is a basic explanation of the component parts which constitute the studio, then you are bound to realise the exciting implications which it has for music in the future — and by that I do not just mean avant-garde or experimental, which is how electronic music is frequently categorised, but music as a whole. As to why this studio must be housed in a garden shed, why it is up to a private individual to develop and finance it, why the lack of commercial or national sponsorship (while the IRCAM project in France attracts ex-EMS teamsters), why the studio has been 'given' away; these questions really come after an understanding of how this pioneering studio works - hopefully this article will provoke some fruitful public discussion and action in this field. Meanwhile Peter Zinovieff continues to work as a seemingly insular 'unsung hero'. On the verge of finally achieving some long-overdue program updating, he found time to give me a guided tour of the equipment, and this forms the basis of my explanation of how the studio works.

Apart from the main studio with computer-controlled equipment in the shed, there's also a large listening and recording room in the house; a 4-track Ampex, plus four Quad amps and four Lockwood monitors share this area with the most amazing selection of odd and unusual musical instruments — marimbas, log drums, electric violin, things that you hit and things that you stroke — there's also a grand piano for more conventional music making. This is a very important and useful part of the studio, as a lot of the computer sounds are not synthesised from scratch, they in fact start life as 'real' sounds, which are then analysed and distorted to match the composer's ideas by the computer-controlled devices.

This method of operation is illustrated by the most recent work that has been done at the studio, which was music for the National Theatre play Brandt, composed by Harry Birtwistle. This was in two parts: the first, a Bach chorale, for which the score was analysed into 'numbers' so that the computer could then play the piece; with the major advantage that the music could be dynamically distorted as it was interpreted. As a counterpart to this, there was a recording of people singing a psalm in a church on the west coast of Scotland. This was recorded on the studio's portable Nagra tape machine, and then fed into the computer, which performed variations and filtering on the basic subject material. Peter Zinovieff seems to be very interested in using this form of transformation, and computer programs (VOCAB) have been designed to interpret and transform this type of input. No doubt this interest has paid off in other respects as well — the analogue Vocoders, for instance, designed by Tim Orr, are now a regular EMS product. I think this sort of interchange is a major part of the computer studio's work, aside from the music and development that comes from it directly. In the same way that NASA's space programme gave birth to a whole host of 'spin offs', Peter Zinovieff's computer programs have doubtless highlighted the possibilities inherent in certain approaches to music synthesis. Just as you can't land on the moon in a Teflon-coated saucepan, neither can you analyse the score of a Bach chorale with EMS's new Universal Sequencer — but you can load it from a touch keyboard with up to 256 notes in real time or single step, and you can replay them in tempo to an external rhythm on any make of voltage controlled synth - simple stuff maybe, but cheaper and far better than its competitors (which look positively antiquated by comparison), and a very useful tool for the electronic music (em) composer.

Back to moonbase... and to the computer studio in the garden. This is divided into two parts; a reasonably quiet area with programming instruments, and a noisy area with the bulk of the equipment — noisy because all the computer racks and so forth have cooling fans which would not be out of place in Battersea power station. Obviously this problem could be surmounted in a proper location with equipment isolation and soundproofing, but with the growth rate and major investment of time and money in hardware this has clearly not been possible. One day it would be lovely to see the studio in this type of environment, though, with proper location of the monitoring loudspeakers and so forth and a link-up with a well-designed conventional control room and recording studio. One half of the present 'noisy area' houses the sort of equipment line-up that one would expect to find in a conventional 'league division three' studio: two old Revoxes, the Nagra, Ampex 4-track with Dolby A, Scully 16-track, Electrosonic 16-input console ('the weakest link in the chain'), JBL 4311 monitors, plus an EMT plate, Astronic graphic equalisers and an Audio & Design compressor/limiter. The adjacent wall leads towards the computer equipment — both spiritually and physically. Next in line by the mixer is the EMS Vocoder, and then the largest of the conventional voltage controlled synthesisers that EMS make, the Synthi 100.

Despite the sophistication of the Synthi 100 as an instrument, it is still rather a cumbersome sound-producing device when compared to the computer-controlled equipment in this studio. Currently it is mainly used as a treatment centre for sounds, sending them off to the EMT echo plate or a limiter, and then passing them through an envelope generator or filter. 'It's as simple or as complex a synthesiser as you want, and it's extremely nice for a studio like this because no patch-cords are necessary. Some of the sliders and joysticks on this one can be remotely radio controlled, it also has some good measuring things, a nice little memory, an oscilloscope, a frequency meter and a DVM.' The lower part of the Synthi 100 contains several 19in rack panels, in which are mounted the conventional studio 'outboard equipment' such as the 9-band Astronic graphics. Also housed in this area are several EMS modules, octave filter bank, pitch-to-voltage converter, a prototype EMS digital oscillator, plus a 'really marvellous' old Wavetek VCO which originally cost £600. Sixty of the voltage-controlled parameters on the Synthi 100 — two reverb mixes, five time parameters on three envelope shapers, filter frequencies on eight filters, frequency controls on all 12 oscillators, eight VCAs, three slew limiters and the sequencer 'load' initiate — are all interfaced so that they may be controlled by the computer or the Computer Synthi.

The Computer Synthi is situated next to the Synthi 100, located at one end of the 'computer rack' wall. The Computer Synthi was designed by Peter Eastty and is an actual EMS 'product' — apart from the one in the studio, Glasgow University and Aura Films in Paris also have one each. The basic function of the Computer Synthi is to interface with any voltage controlled synthesiser, such as the Vocoder or Synthi 100 (which it matches in case styling), or a full size DEC PDP-8 computer. The link is via 24 analogue-to-digital converters and 24 digital-to-analogue converters, which are connected to specific VC devices via two 64 x 20 patch-panels.

In its standard form, the Computer Synthi has two 'onboard' cassette tape drives for storage, a 64 push-button program control panel, an alphanumeric display, realtime clock, and 16 faders which give variable control of parameters selected at the patch panels. In this self-contained 'basic' format the Computer Synthi can be used as a 'real time editing' multi-level sequencer; interfaced for example, with the Vocoder, it could store various combinations of the 22 different voltages which the Vocoder produces to correspond with the energies in the 22 different frequency bands of its analogue filter bank: these voltages could then be orchestrated by the composer, and given out again by the computer in whatever sequence and with whatever dynamic modifications the composer wanted to make, thus controlling the Vocoder. The great advantage of the Computer Synthi is that, with the addition of a printing terminal, plus DECtape or a disc for extra memory, it can then run the very sophisticated MUSYS language, developed by Peter and co over many years, which then enables really involved composition using computational facilities. 'It's an expensive piece of equipment, it cost about £20 000, and many people would say that it's not the way to do it, that you should have a microprocessor and a floppy disc; on the other hand, this is a PDP-8 and we can connect it to our main computer system, all the program development can be done on the main computer system, and it uses the same sort of technology that we've used on our other machines. The case and the patchboards cost much more than the actual D-to-A converters - even the computer in it is only a thousand quid's worth or so.'

One of the possible uses of the Computer Synthi which occurred to me while examining it was that, in conjunction with an analogue VC synth like the Synthi 100, it could be used to store a catalogue of complex oscillator tunings and envelope/filter sounds, which could then be assigned at will to play different parts of an edited composition sequence programmed by the composer. 'In fact, we haven't used it like that,' Peter told me, 'because no-one has come here yet with a composition which requires that approach. Also it really isn't completely de-bugged yet, and the one in Paris has just blown up so we've got some of the boards back. I wouldn't want people to come here and just use the Computer Synthi as a sequencer, because it's a bit too simple: you can get stuck if you want to do some more complicated things because it's too slow, whereas if you hadn't had a good computer system like us then it would be very magical, I should think.'

Along the wall next to the Computer Synthi are the racks which house the main computer system, plus associated equipment and specially developed EMS devices. The computer itself is a 24-thousand-word DEC PDP-8E with fast arithmetic, a disc and DECtape for storage, and it works with an operating system known as OS8. Peter Zinovieff: 'All of our programs fit into this operating system, so that in the end we have a program developed of several hundred thousand words - it's an enormous bundle of programs - all of which now work. Some of the programs do things like decide mathematical functions to do with music, some control devices, and some operate just one piece of equipment, like the light display. They all add together into this one system that we call MUSYS, and when you are 'in' MUSYS, you can then have access to any of the other facilities.'

Above the computer is a new machine known as a Fast Fourier Transform Analyser, or FFT for short; this device can be best explained as the equivalent to 500 bandpass filters at linear intervals, but distributed over any frequency range of the spectrum. When a sound is processed by the FFT, it is first converted into numbers by an analogue-to-digital converter, a mathematical formula analysis is made of the numbers, and the 'partials' of the sound, in terms of sine waves, are then given out at 500 points of the spectrum. This is obviously a very powerful piece of machinery to have, especially when using 'real' sounds (such as the church choir) as the basis for computerised music. 'With this, one should be able to recreate any sort of sound. I reckon that the future is to have an inverse of the FFT machine to produce the sounds, and I think this is what we'll have next year, and not use oscillators at all. But that's only because we have programs that are able to deal with 'frames' of data just like you get from an FFT machine.' This is a very spectacular proposal — that the elegance and versatility of computer additive synthesis could be obtained, but without the awkwardness and slow processing time; FFT can perform an analysis in less than 30 milliseconds, and, as Peter pointed out, EMS have already spent a lot of time in research and development of programs which permit rapid and intelligent compositional dialogue with computer music systems.

One rack along from the computer and FFT are several of the specialised EMS devices. The 'Analytical Engine' is, as its name suggests, another analysing filter bank, although the 128 filters here are much more variable: they can be programmed to have different 'window' functions, any bandwidth, any 'Q', and they can be set at any frequency with an accuracy of 0-25 Hz. 'That's probably the most advanced analyser there is anywhere in the world.'

(To be continued).