In the final part of our two-part Retrozone, Gordon Reid charts EMS's further achievements, the reasons for their decline in the '70s and the company's phoenix-like rebirth in the '90s.
In Part 1 I described the genesis of EMS, and the arcane architecture of the VCS3. This instrument was a great success at the time of its launch in 1969, and EMS briefly vied with Moog and ARP at the very top of the synthesizer tree. Indeed, it's interesting to speculate about how today's synthesizers might look and feel had the company gone on to release the VCS4 and the Synthi KB1. But, as we now know, the Minimoog later set the pattern for all things to come.
Neither the VCS4 nor the Synthi KB1 made it into production and EMS's next synthesizer was the huge, unwieldy, and almost unusable Synthi 100. Contrary to the information on some web sites (which are simply wrong), this huge floor-standing unit was released in January 1971, and looked like one of the computers in the original series of Star Trek. (You could order the Synthi 100 itself as three large rack-mount modules, but it's the floor-standing version that has survived in memory.)
The Synthi 100 used the same technology as the VCS3, but incorporated 12 VCOs, a pair of noise generators, three ring modulators, four low-pass filters, four high-pass filters, plus 39 other modules, and an oscilloscope! You played it using a pair of five-octave velocity sensitive keyboards but there was also a six-track, 256-step digital sequencer — a radical innovation then.
At £6,500, it was obvious that the Synthi 100 was never going to be a musician's instrument, and almost all of the 30 or so produced ended up in universities or in broadcasters' in-house audio workshops. The most famous of these was the one installed at the BBC Radiophonic workshop in Delaware Road, London and, in honour of this, the Synthi 100 also became known as 'The Delaware'. (If you want to hear 'The Delaware' you need only listen to the original radio series of The Hitchhiker's Guide To The Galaxy, or to any of the Jon Pertwee episodes of Dr Who.) The same year also saw the development of the one-off Sequencer 32, the Sequencer 64, and the Sequencer 128 — all of which evolved into the Sequencer 256, a stand-alone version of the six-track sequencer in the Synthi 100. This was no small box but, with its five-octave keyboard, 42-bit storage, internal and external clocks, forwards and backwards play, and editing capabilities, it was incredibly advanced in 1971.
You could later buy some of the Synthi's other modules as separate products designed to complement the VCS3. These included the pitch/voltage converter (possibly the first of its kind), a triple slew generator, a filter bank, and a random signal generator, none of which existed in the VCS3 itself. But neither the Synthi 100 nor its spin-offs proved to be EMS's crowning glory. That honour belongs to two direct descendents of the VCS3. Released in 1971, the first was the Synthi A, or 'Portabella', and the second was 1972's Synthi AKS.
Pink Floyd's Dark Side Of The Moon is famous for its extensive use of the VCS3, and 'On The Run' is perhaps the classic VCS3 track... but there is no VCS3 on it! Alan Parsons (the engineer on the album) later revealed that the track was recorded using a Synthi AKS. It was, as he put it, the "hot synth of the year".
Because of the lack of synchronisation technology at the time, the track (except for the spun-in sound effects) was recorded live. The sequence was laboriously programmed using the AKS's membrane keyboard, and then played back at high speed to produce the effect that you hear. Considering the complexity of the sequence, the amount of sound modulation going on, and the hi-hat sound that is also being produced by the synth, it's remarkable that everything worked as well as it did. More so when you realise that the same rigmarole was required before every concert — there were no stage sequencers in 1972!
Designed by David Cockerell, the Synthi A was functionally identical to the VCS3, the difference being that it was built into a Spartanite attaché case. At just £198, it proved to be a winner, popping up in numerous colleges and music studios. Nevertheless, its success was relatively short lived, because the AKS added an unplayable membrane keyboard and a sequencer in the lid. At £420, the AKS cost more than double the 'A', but this was the machine that everybody wanted. Soon the user-list began to read like a Who's Who? of the music industry. It included electronic artists such as Tangerine Dream and (somewhat later) Jean-Michel Jarre, prog-rockers King Crimson, Pink Floyd, Yes, and more orthodox but far-sighted artists such as Roxy Music, David Bowie, and Led Zeppelin. In fact, the second synthesizer I ever used was a Synthi AKS, owned by the music department of my local sixth-form college. I spent hours battling to get tunes out of it, and still more trying to master the membrane keyboard. I was greatly disappointed when I failed, and it wasn't until some years later that I discovered that everybody else had had the same experience!
Also in 1972, the DK2 was released — this was a duophonic version of the DK1 which you couldn't use duophonically, because when you depressed a second key, the second CV dragged the first one down. In addition, there was the KS Sequencer, which was, in essence, a stand-alone version of the sequencer and membrane keyboard from the Synthi AKS.
Looking back, it's obvious that 1972 marked the zenith of EMS's achievements. In 1973, the company released the Sound Freak, a guitar synthesizer that came as a stand-mounted console plus two CV pedal controllers. Later renamed the Synthi Hi-Fli, this proved to be Cockerell's first failure. It wasn't short of facilities — it offered distortion, ring modulation, a basic envelope generator, modulation, and various filter treatments — but it never caught on (it's a shame that ARP didn't learn anything from this, because the ARP Avatar guitar synthesizer was a significant factor in that company's death eight years later).
On the positive side, 1973 also saw the release of the VCS3 MkII. This incorporated a stronger power supply and slightly improved oscillator stability, as well as the 'Prestopatch' socket. This socket allowed you to use 16 x 16 plugs (custom built to customers' specifications) that mimicked the patch matrix, providing primitive patch memories for the routing of a sound. Of course, you still had to set all the controls by hand, but it simplified some of the sound creation and eliminated one source of potential cock-ups if you tried to use a VCS3 on stage. Unfortunately for EMS, technical improvements like this did not lead to increased sales. Throughout the heyday of EMS, its driving forces and inspirations had been Peter Zinovieff and David Cockerell. But by 1974, Zinovieff's interest was waning. He was computer-literate in an era when very few musicians were and, throughout the early '70s, his ideas remained way ahead of their time. Furthermore, he had used EMS as a platform to finance his research in his MUSYS studio, and this was still to lead to many developments in new ways of using computers to shape and reproduce sounds. But conventional synthesis was no longer of real interest to him, and he had started to experiment with various forms of communications technology and speech encoding.
At the same time, EMS's research projects were becoming ever more ambitious, and much of the company's revenue was being diverted into the development of a huge digital synthesizer (the 'Digital Oscillator Bank') that was to have offered precise control over the waveforms and envelopes of its 192 oscillators! Yet another monster, Peter Eastty's 'Analytical Engine' incorporated 128 digital filters that produced the raw data used to drive the DOB. Eastty had one fully-functional version of this in his studio but, when I discussed it with him in 1998, he confided that he had designed it as research tool and had never really seen it as a commercially viable product.
On the commercial side, Cockerell had continued to develop his VCS3 technology, producing two 'Matrix' products (one with keyboard sockets, one without) that allowed you to connect the Prestopatch sockets of two Synthi As or two VCS3 MkIIs. He had also developed a new version of the VCS3/Synthi A.
- David Cockerell left EMS in 1972 to join Electro-Harmonix, the American manufacturer of effects units and miniature synthesizers. He spent a brief time in 1976 at IRCAM in Paris, but left to return to Electro-Harmonix. More recently, he was instrumental in designing Akai's hugely successful 'S' range of samplers.
- Peter Eastty left EMS in 1977, also moved to IRCAM, and then spent some time at SSL, perhaps the world's most successful manufacturer of large-scale mixing consoles. He now works for Sony Broadcast. As senior scientist at the company's Oxford research labs, he has helped develop DSD and Super Audio CD (SACD), the new consumer CD format that Sony hope will displace the — as yet unreleased — DVD Audio format.
- Apart from his work at EMS, Tim Orr is best known for his Powertran synthesizer kits (as published in Electronics Today International magazine) such as the Transcendent 2000. I recently gave a lecture on 'Analogue Synthesis, Past and Future' to the British section of the Audio Engineering Society and Tim was in the audience, which was terrifying!
- When Graham Hinton left EMS in 1979, he briefly left the music industry, joining computer manufacturer RML (Research Machines Ltd) to work on local area networks. More recently, he returned to audio, helping design products for HH Electronics and SSL.
Named the Synthi P (the 'P' stood for Professional) the new instrument offered all the features of the 'A', but incorporated improved filters and envelopes, plus a heater on the oscillator chips to improve pitch stability. There were many other improvements. For example, Cockerell added pulse width modulation, a dedicated filter on oscillator one, oscillator sync between oscillators two and three, slaving of oscillator three's pitch to one of the shapers, true ADSR contours, filter tracking, and portamento. The 'P' also sported a new mechanical keyboard, somewhat like a DK2 but with a KS sequencer incorporated within it. Unfortunately, the people who played the 'P'-- and there aren't many of them — say that, whatever it gained in performance and reliability over its predecessors, it lost in character.
Whatever the future of the Synthi P might have been, there were only three prototypes made. One — sometimes called the Synthi Mk3 — was housed in a Synthi A case, the other two appeared in small flight cases. None saw the (commercial) light of day, although at least two of the prototypes still exist, and now reside in the USA. Likewise, Cockerell's remarkable Speech Synthesizer (which allowed you to draw the formants with a pen) never proceeded past the prototype — possibly because Cockerell had left the company in 1972!
So, in December 1974, EMS released its first product that did not have the Cockerell name attached. Richard Monkhouse's Spectre (or 'Spectron Video Synthesizer') was a remarkable video synthesizer that combined digital and analogue technology, and sported yet another of EMS's trademark patch matrices. It could produce its own shapes and colours, and you could use it to modify existing video signals. It even allowed you to modify the video image using an audio signal. Unfortunately, and like so many of EMS's designs, the Spectron was too innovative and too complex for its time. Consequently, only 15 were built.
Yet despite these diversions, flops, and a general lack of direction, EMS continued to bring viable products to the market. Much of this was thanks to a young designer named Tim Orr, whose first release was the Synthi E, yet another derivation of the VCS3, slightly cut down for educational use (hence 'E') and again packaged in a briefcase. He complemented this with the Synthi DKE, a three-octave mechanical keyboard that rendered the horrible membrane keyboards of the 'A' and AKS unnecessary.
Another Orr design was the 'QUEG' — the QUadraphonic Effects Generators — a four-channel surround mixer and effects generator. Neither of these proved to be huge commercial successes, but neither were they horrendous flops, so EMS continued to... well, if not prosper, at least survive.
Unfortunately, Orr's designs could not halt the slide of the company, largely because the other developers' projects were, to be polite, less than commercially oriented. One of these projects was Peter Eastty's Computer Synthi — a digital controller for the Synthi 100. In a year when a three-bedroom suburban semi cost less than £10,000, this had a projected end-user price of £25,000. With a total market of virtually zero (ie. only those studios that already owned a Synthi 100), it was never going to be a resounding success. Sure it was technologically remarkable... it incorporated a PDP8/M mini-computer, 24 analogue-to-digital and 24 digital-to-analogue converters, cassette decks, and CV controllers, but who needed it? EMS built just three.
In 1976, following the failed launch of Zinovieff's 'International Voice Movement' telecommunications company, EMS left London and relocated its development team and music studio to The Priory in Great Milton, Oxfordshire. But by this time, the company was in terminal decline. Yamaha had already demonstrated the future direction of synthesizers with its GX1, and the hugely successful CS80 polysynth had just appeared. The Oberheim 4-Voice was also available, the Polymoog had shown a different flavour of polyphonic synthesis, and the Prophet 5 and the Oberheim OBX were just a few months in the future.
In this climate, the unstable and temperamental VCS3 and its various descendants were no longer attractive. Yet EMS was still producing innovative products and, in 1976, it released Tim Orr's now classic Vocoder 5000 (which, perhaps not entirely coincidentally, cost £5,000). This incorporated no fewer than 22 band-pass filters and envelope followers, together with oscillators, noise sources, a frequency shifter, a spectrum analyser, and comprehensive interfacing. It sounded good too! Indeed, in the July 1977 issue of Studio Sound (a magazine not noted for hyperbole) reviewer Nik Condron stated that, "It is my belief that in terms of all kinds of music synthesis, this machine will be the forerunner of the final stage of musical technological development — and perhaps it is at this time that the question should be asked: Where do we go from here?"
EMS also produced a cheaper and more limited Vocoder which, nonetheless, still sounds good today — Tim Orr's Vocoder 2000 — and his 'Universal Sequencer', an expanded KS Sequencer with standard CV and Gate interfaces. But, on the other hand, there was the PolySynthi...
There have been many keyboards that used the so-called 'Paraphonic' keyboard technology: divide-down oscillators passed through a single filter and a single envelope. Of these, a couple enjoyed moderate success — most notably the Roland RS505 and ARP Omni string ensembles --but the class also included numerous, horrible, 'Polyphonic Ensembles'. By 1978, it was a discredited approach to polyphonic synthesis, so why designer Graham Hinton adopted this for EMS's much-needed polysynth is a mystery. Perhaps financial constraints were a major factor? The company had never before demonstrated such a lack of technical innovation. Despite costing a mint to develop, and notwithstanding the Electro-Harmonix MemoryMan, which was incorporated to thicken up the sound with grungy chorus/delay effects, the PolySynthi sounded thin and uninspiring. Dave Cockerell was at that time working for Electro-Harmonix helping to design the Electric Mistress and the MemoryMan. When EMS tooled up to produce the PolySynthi, he brought the effects boards back to the UK in his suitcase! With fewer than 30 ever built, the PolySynthi was another EMS disaster.
Unfortunately, Hinton's other projects for EMS were equally disastrous. These included 1978's VCS3 upgrade cards (which were to offer improved VCOs, better VCAs, ring modulation and reverb), and a polyphonic sequencer co-designed with Peter Zinovieff. Apparently, both cost a great deal of money to develop, but neither made it past the prototype stage.
Synthesizers are not truly polyphonic simply because they can play many notes simultaneously. To earn this description each note has to be shaped individually, and without deference to any notes that may already be sounding. This is accomplished in one of two ways. The first, as used in the wonderful Korg PS3200, is to have one or more envelope generators and filters for every note on the keyboard. The second, used by almost every analogue polysynth from the Prophet 5 onwards, is to restrict the number of notes that will sound simultaneously, and assign a limited number of envelopes and filters as the notes are played. Unfortunately, the PolySynthi did neither. While it allowed you to press every note on its four-octave keyboard simultaneously, its single envelope and filter meant that it couldn't correctly articulate the notes that you played.
In 1979, EMS's financial problems became insurmountable when one of its European distributors refused to pay its debts. Liquidators were called in, and they sold off the equipment in the studio, piece by piece. EMS itself was sold to Datanomics, a company based in Wareham, Dorset. EMS's products had always been built in Dorset, and a number of the people formerly involved in manufacturing VCS3s and their descendants now worked for Datanomics, so it's possible that they saw a future for the company where others could not. Whatever their motives, from 1980 to 1984 they kept the EMS name alive, building occasional VCS3s, Synthi AKSs and Vocoders.
Datanomics also provided the finance to develop two new synthesizers. The first of these was yet another monophonic descendent of the VCS3. This embraced digital technology with a Z80 microprocessor handling the housekeeping, and replaced the VCS3's pin-matrix with a huge array of 400 programmable membrane switches and LEDs. It used Curtis chips for its three VCOs, and also sported a second envelope generator, high- low- and band-pass filters, and an analogue delay line. Yet another EMS disaster, the DataSynth would have been as large and unwieldy as a PolySynthi, and the prototype proved to be horrendously unreliable. With a projected price of £3,500, it's hardly surprising that it never made it into production. The second was a new Synthi 100, also based on Curtis chips, but only one of these was ever built (it ended up in a college music studio in Spain).
Following Datanomics' failure to revitalise the company, its next owner was Edward Williams, a composer who owned an EMS-equipped studio in Bristol and who had written music for TV series including Life On Earth and other natural history programmes. Williams was interested in combining music, technology and dance so, in 1985, he recruited Richard Monkhouse and, together with Robin Wood (who had stuck with EMS throughout all its highs and lows) they developed a new product.
Originally intended as a musical generator designed to complement avant-garde dance, the new device used ultrasonic range-finding technology to turn distances into musical information. It took another three years to bring a MIDI-equipped version to the market but, in 1988, the 'Soundbeam' became EMS's first commercially viable product for more than a decade. Designed for performance artists, the Soundbeam's ability to convert distance to voltage to MIDI allowed players to convert physical movements into sound, much like a Theremin. But Soundbeams also allow you to use tiny finger movements or even your eyelids to control MIDI modules, thus making them suitable for use in special needs schools. Indeed, physical and occupational therapists have noted that the Soundbeam gives some severely disabled people their first opportunity to control their environment with the minimal movement available to them. As such, it has become something far more important than just another type of synthesizer controller.
In 1995, Robin Wood finally acquired all the rights to EMS, including the rights to manufacture the Soundbeam — this was 25 years after he started working for the company. For many years he had continued to supply EMS Vocoders, together with parts, manuals, and servicing for VCS3s and the various Synthis. Now he could again start to build VCS3s and Synthi As, selling them for £1,800 and £1,600 respectively.
Wood manufactures his new models to the original specifications although, if you require, he will also build special models that incorporate the modifications described in part one of this Retrozone feature. But be prepared for a considerable wait... there is a long queue for these hand-built classics.
Soundbeam itself became a separate company — The Soundbeam Project-- based in Bristol. It continues to develop the Soundbeam, and now boasts more than a thousand users worldwide. In 1999, it released Soundbeam 2, an enhanced Soundbeam with memories, extra sound generating functions (such as programmable pitch sequences) and a switchbox that lets you connect other controllers such as joysticks. It also became speed sensitive as well as distance sensitive, opening up many more performance possibilities. With two sensors, Soundbeam 2 even allows players to change parameters and MIDI programs while playing sounds.
When I showed an early draft of this article to Peter Eastty, he was rather concerned that I had got the balance wrong. He suggested that I had made rather too much of EMS's failures without devoting enough text to its successes. So let's make sure that the following is clear. There was a time when EMS stood shoulder-to-shoulder with Moog and ARP. Indeed, in many ways, EMS was more advanced than either of these, and it's very difficult in 2000 to appreciate just how radical Zinovieff's and Cockerell's pioneering ideas and developments had been. But, in contrast to the genuine 'firsts' such as the VCS3 and DK1, products such as the Computer Synthi were spectacularly unsuccessful. As for the PolySynthi... well, I still find that chapter of the EMS story completely inexplicable.
Is this a balanced view of the highs and lows of EMS? I hope so, and I hope that I have shown how the company could grow so quickly from almost nothing in 1969 to world leadership in 1972, and then plunged almost as quickly into bankruptcy by 1979.
So, what can we learn from the EMS story? Maybe it's something that we already know. The Brits are often ahead of their time but, in the music industry as in so many others, they are unable to exploit their ideas in the way that (for example) the Japanese do. At EMS, as elsewhere, the dreamers and pioneers returned to earth with a hard commercial thump, and it was those who copied and later refined their ideas that eventually went on to rule the synthesizer world.
But the spirit of EMS has refused to die. The UK still has some — albeit small and very specialised — synthesizer manufacturers. One of these is EMS itself, and we should all be very grateful to Robin Wood for keeping the brand alive. Another is Analogue Systems, whose designer, Steve Gay — you guessed it — once worked for EMS. Maybe, just maybe, somebody in the year 2030 will be able to write a retro about the second 30 years of EMS.
I would like to thank Robin Wood, Peter Eastty and Steve Gay for all their assistance in researching these articles.
In Germany, there is a company called EMS Rehberg. This is owned by long-time EMS fanatic and user Ludwig Rehberg who, I understand, created some of the Synthi sounds for Dark Side Of The Moon and Oxygene. The company has distributed EMS and Soundbeam products for many years, but also manufactures products using the EMS-Rehberg name. Tim Orr designed at least two of these. The Analyse-Filterbank B1 was a 10-band graphic equaliser sold, primarily, to German broadcasters. Far more interesting than this, the Synthi-Logik was an improved Synthi E with several enhancements including a mechanical keyboard, portamento and a true ADSR contour generator.
EMS Rehberg Inkenstrasse 4, D-71254 Ditzingen, Germany.
+49 (0)7152 53273.
EMS, Trendeal Vean Barn, Ladock, Truro, Cornwall TR2 4NW.
+44 (0)1726 883265
Graham Hinton now provides a great deal of excellent EMS and VCS3 information. A full list of modifications together with a great deal of other information, exists on his web site.