From leading the pack with the Minimoog, Moog fell badly from favour with the ill-fated Polymoog. What went wrong, and was the company's first polysynth really a mistake?

IT'S NOT THE easiest task in the world to write a retrospective of an instrument which Julian Colbeck dismissed with such uncharacteristic ferocity in the first edition of Keyfax (before deleting it altogether from Keyfax 2). Let's face it, no-one's had a good word for the Polymoog in the last ten years - but I've got a few. I've just spent a few weeks renewing my acquaintance with the beast, and I positively like it.

In the mid-'70s, there were monophonic synths (with the Minimoog the undoubted leader of the pack), the Clavinet, electric pianos and organs, and the odd string synth. What there wasn't was a proper polyphonic synthesiser, one with the degree of control you could get from a monosynth. The race was on to produce the first commercial polysynth, and Moog had started putting money into the task as early as the early-'70s. Their chief designer, Dr David Luce, asked for and received feedback from a who's who of then current keyboard luminaries, including Herbie Hancock, Stevie Wonder, Chick Corea, Jan Hammer, Patrick Moraz, Larry Fast, Roger Powell and Keith Emerson (plus Michael Tillson Thomas from the world of classical music). Moog incorporated their suggestions wherever possible into their design, producing the very first Polymoogs in the winter of 1975/76, and making many modifications before production proper started (I believe) sometime in 1977. The effort reputedly cost them half-a-million dollars, which is probably peanuts to, say, Yamaha's R&D department nowadays, but was an extraordinary amount of money then. The question was, had they spent the money on the right things? With the benefit of hindsight, the answer has to be yes - and no. Some things they did absolutely right but others proved commercial liabilities which meant the Polymoog had only a very short time at the top of the heap.

The Polymoog is rather an oddball, and one which has certain similarities to the Yamaha CS80. Both sets of designers went to great lengths with keyboard control, and their instruments were far superior in this respect to the Prophet and Oberheim synths that came along afterwards and took away most of their sales potential. The Polymoog's keyboard isn't up to the wonderful feel and polyphonic aftertouch of the CS80, but it still feels good, even today. It's a strange length - 71 notes, E-D - which is slightly shorter than a JX10, but much better than your standard five octaves. (I believe early prototypes ran F-E, and were thus 72-note keyboards; they also had more spindly knobs and more sticky-up buttons.)

Couple this with the ability to split it into a (nearly) two-octave lower and four-octave upper section, and give it touch sensitivity and a very "un-plastic" feel, and you have the potential for a superior performance instrument. There aren't any pitch or mod wheels, admittedly - only a rather small and tacky ribbon controller placed awkwardly dead centre - but the Polymoog makes up for that with a vast array of foot pedal control sockets. It always seemed an anomaly that so many otherwise decent synths didn't give you the chance to use the dangly things at the end of your legs to alter sounds, and Moog obviously thought so too. Hence the Polypedal, a rather expensive unit with controls for filter, pitch, volume, sustain, triggering mode, modulation, and the ability to switch in and out an external synth. Even if you didn't have a Polypedal, you could do the same jobs with a shopfull of other pedals and switches but you should be aware that the four switch sockets are non-standard jacks (marginally smaller than the standard quarter-inch) to prevent you mixing audio signal and control signal paths.

Talking of interconnecting, the other thing to remember is that we're talking long, long before MIDI. Consequently, the only synths you could hook up to the Polymoog (and play with the highest note on either the lower or the upper section) were other Moogs like the Minimoog, Micromoog or Multimoog, which used Moog's S-trigger system rather than the later standard V-trigger.

Back on the front panel Moog boasted that they'd done the best ergonomics job ever, with all controls within four inches of the keyboard. It's true, and if you've ever lunged at controls on an Oberheim or a Prophet 10, you'll appreciate this setup. All controls are on a steeply-angled panel at your fingertips. Additionally, the Polymoog has a good flat top to it which is ideal for perching another keyboard (or a few pints of lager) on. The controls incidentally, especially the sliders, are good and solid. My Polymoog must be getting on for 15 years old, and the worn numbering on the preset buttons testifies to a good deal of use, but all the sliders are still smooth and (largely) completely noise-free.

So the controls are robust and within easy reach of the keyboard - but are they useful? They are, but some more so than others. It was a good idea to include some sort of relative keyboard balance control, especially with the Polymoog's capacity to split the bottom two octaves from the top four. But giving you three double-octave balance controls is silly. The top four octaves are bound to have the same sound, so any discrepancy in their relative balance just makes a run or chord crossing the change-over point sound like sloppy dynamics. More useful are a plethora of other controls by which you can differentiate the lower section of the split from the upper, including separate waveshapes - either or both - square-wave pulse-width and modulation; relative amounts of sawtooth wave; decay time; and whether the resonator section and VCF affect either or both.

What this means is while you can't set up two completely different patches each side of the split point, you can make them sound surprisingly different. You can have a pizzicato double-bass line down below and a sustained pseudo-horn section above, or any number of other pairings of more or less usefulness. But if you find a mega-patch at your finger tips, out come the pen and paper. Yep, the only thing that's worse - and less reliable - than cassette downloading is human downloading onto what we old timers used to call patch sheets. There's actually something quite satisfying in filling in little black blobs on a genuine Moog patch sheet. What's less satisfying is setting the patch up again a week later and finding it just don't sound quite the same.

"On the Polymoog's front panel moog boasted that they'd done the best ergonomics job ever, with all controls within four inches of the keyboard - it's true."

The Polymoog came just too early to benefit from the microprocessor revolution of the mid- to late-'70s, and consequently has precisely zero memories - worse even than the four miniature sets of faders the CS80 boasted. It also missed out on the other turning point of the era - the idea of having a limited (and thus cost-effective) number of voices assigned by a scanning micro-processor to whatever notes your fingers hit. What Moog did instead was what makes the Polymoog unique, but probably what made it so uneconomic to turn out that I gather they even halted production a couple of years before the Memorymoog superseded it. They designed a chip which included practically everything a synthesiser voice needs - two VCAs, a VCF, and an envelope generator - and then calmly shoe-horned 71 of the little blighters into the Polymoog - one for each note. The virtue of this is that you have absolutely unlimited polyphony - a bit of overkill, but not to be sneezed at if you're using a lot of notes with long decay, because it means that you can just keep on playing them without worrying if the next note you play is going to silence an earlier one mid decay. On top of this, each note is marginally different from its neighbours in subtle ways related to component tolerances and very much like the subtle differences between piano strings.

The downside is that something had to be simplified if the Polymoog was going to fit into a normal size case, be reasonably affordable and avoid horrendous tuning problems. The solution Moog adopted caused a lot of controversy: they borrowed organ technology, and used a very high master oscillator for each of the 12 notes, dividing down to get the lower octaves. This led to accusations that the Polymoog was a glorified organ. It wasn't really true, because for one thing the touch sensitivity made it very un-organ-like and for another, the oscillators were phase-locked to VCOs which enabled them to be modulated in a way an organ couldn't be. But what it did was to set up some bad publicity which, when added to another quite reasonable popular prejudice of the time, was enough to shorten its commercial life considerably. The purists sneered - the Polymoog wasn't just a glorified organ, it was one of those dreaded preset synthesisers as well. Nowadays, you wouldn't get far selling a synth (or sampler) without an impressive load of sounds to go with it; but remember, this was in the days before the early Prophet 5's magnificent (if totally unreliable) 40 memories began our drift towards the "factory preset" syndrome.

It was all too much for a lot of people brought up on Moog modulars, or on the elegant little Minimoog. What made it worse was that, although some very acceptable traditional Moog sounds could be coaxed out of the Polymoog, they didn't spill out of it as they had from the Minimoog. The classic bass sounds in particular just couldn't be reproduced, no matter how hard you tried.

Of course the lack of a brilliantly punchy bass sound isn't the end of the world for a polyphonic synth, particularly when it was a reasonable assumption that most potential buyers already owned a Minimoog or an Odyssey. (Of those two monophonic princes, the Polymoog sounds to me more like the Odyssey. Don't ask me why.) What you got in the Polymoog was eight preset instruments - or modes, the designers called them, aiming that you should use them as starting points. String is a bit thin but OK as it stands; Piano was good for its time and sounds like a nondescript touch-sensitive electric piano; Organ is pretty poor (interesting when you think of the criticisms); Harpsichord is thin and unappealing (though apparently meatier in late models - s/n above 3000); Funk is not exactly superbad but not bad either; Clav is reasonable; Vibes is hollow and weak; and Brass is good, though not up to Oberheim standards.

There are, the eagle-eyed will have noticed, two more buttons I haven't mentioned. One is called Var, and enables you to free up all the controls to put your own sound in. The only trouble is, it doesn't stay in memory if you switch off the Polymoog, or even if you go to one of the presets and move a control or two. The second (.) button is a bit of a disappointment as well. Its function is simply to transfer the settings you've come up with as an adaptation to any of the original presets. This can be quite pleasing, but it's not what you would call a usable memory.

"People presently pay a couple-of-hundred quid for a Moog parametric - for the same sort of money you can get this and a Polymoog thrown in."

If the presets were the whole story of the Polymoog, then I'd agree that its consignment to oblivion was justified. (There was actually a purely preset Polymoog, brought onto the market a couple of years after the original Polymoog, and at a considerably more reasonable price. It had around a third of its big brother's controls, 14 average to good presets and the same quality keyboard. There's some confusion here, because Moog called this the Polymoog Keyboard, and renamed the original one the Polymoog Synthesiser. But the original Polymoogs were actually called Polymoog Keyboards. Put more simply, an early Polymoog Keyboard (like the one illustrated) is fully controllable. A late Polymoog Keyboard is not.) It's the controllability of the full Polymoog which redeems it, and, I reckon, makes it a worthwhile instrument even today. For a start, you've got a good (and easily overdriven) filtering section, based round Moog's wonderful 24dB/octave system, and with variable keyboard scaling (eat your heart out, OBXa owners) and even sample-and-hold control of filter cutoff. Then you've got three-band parametric EQ, switchable to low-, band- or high-pass filtering. This actually comes in the audio output chain, rather than in the guts of the synthesis, but what it does is completely transform sounds with the push of a slider or two, making a weedy reed warm and chunky, or a broad string wash into something ethereal. You do have to be a little careful that pushing the gain to its maximum may cause distortion, but even that can be a positive feature, giving warmth and power - particularly in beefing up the poor organ sound into a vastly overdriven Hammond lead sound.

Last but not least, there's one more unique feature the Polymoog has, that sets it apart from any synths except the big modular systems. This is its five separate outputs. I know that perfectly ordinary synths like the JX10 and multitimbral modules like the D110 have loads of outputs but hold on - the outputs on the Polymoog emanate from different sections of the one basic sound source.

Whatever you play comes out of four different outputs: one filtered by the settings of a particular preset; one filtered by the panel-controlled VCF; one controlled by the parametric EQ/filters that Moog call Resonators; and one which is the unadulterated sound of the voice chips. You have a neat five-way mixer for these four and an Aux sound which you can hook up via the socket-laden back panel. Thus, you can use them for separate sends for each section of the keyboard (by setting, say, the VCF to Upper Only and the resonators to Lower Only), but you can also have subtle blends of two or more of them. Whether deliberately or not, they are sometimes out of phase with each other, and this can enhance some of the sync sounds that the keyboard already produces.

For recording, you can put these different sounds through the separate outputs; or for live work you can simply blend them through a mix out (balanced XLR or unbalanced jack). Either way, you can produce subtle changes in tone colour or drastic turnabouts very easily and organically.

A good example of this is to be found in the Piano preset which, with some changes to the VCA, can produce a clean electric guitar impersonation through the resonance output and a dirty one if you whack up the gain. Alternatively you can get loads of different brass or string sounds by varying combinations of output, either as a live performance effect or mixed at your leisure from four tape tracks in the studio.

The microprocessor revolution and MIDI quickly consigned the Polymoog to the out tray. But it's far more pleasant to play than most reasonably-priced modern keyboards, the presets are helpful starting points, and it does have the unique combination of massive polyphony, multiple outputs and its inbuilt three-band parametric EQ. People presently pay a couple-of-hundred quid for a Moog parametric, and for the same sort of money you can get one and all the other bits of a Polymoog thrown in. Hook it up to some digital effects, be prepared to work at getting the feel of the controls and you can get a good variety of really acceptable sounds out of it, from subtly-changing string washes to fat filter brasses to screaming organ histrionics. All this and a little bit of history too.

As well as the famous players who helped in its development, other Polymoog users included Steve Winwood, Giorgio Moroder and, of course, Gary Numan. Not the most currently fashionable group of musicians, agreed, and I can't see the rehabilitation of the Polymoog happening overnight, but it may well come.

Picking up on MT's recent Desert Island Keyboard challenge, if I had to take just one keyboard to a desert island, then the "dreadful" Polymoog (Julian Colbeck) would rate above any number of modern synths, because it's got a decent keyboard and, once you crank up the filter or parametrics, you can get some pretty awesome sounds. Honest.