From: XRJDM@SCFVM.BITNET (Joe McMahon) Subject: Man-Machine Interface (part 1 Date: 2 Jul 90 21:05:38 GMT
A reposting of Metlay's ur-article on the man-machine interface…
The Man-Machine Interface
Well, the response has been remarkably strong, so what the hey: there follows below Part 1 of a detailed, historically and contemporarily relevant, and pretty much irreverent look at the whole process of user interface and electronica, past and present. Structure will be somewhat flaky, but I'll do my best to keep it readable. Also, watch out for smileys |→, as my own twisted opinions have a way of mixing in with the less arguable factual content of the piece. Okay? Here we go!
SYNTHESIS INTERFACING PART 1: THE KEYBOARD
The whole relationship between the synthesizer and the ol' black'n'whites has become so well-established, so historically hidebound, so, well, INBRED, that after a while any attempt to interface man and machine by non-keyboard means has come to be viewed as unnatural and a bit painful, rather like the short-lived pastime of rhino toss. The fact is, the keyboard was NOT by any means the only inevitable controller for the synth, and that in fact a huge number of man-hours have been devoted in recent years to getting around its many shortcomings!
Own up and be honest, people. How many of you would have bothered to learn how to play keyboards if they WEREN'T essential to synth operation? I'll admit it: I learned to love the piano for its own sake YEARS after I'd begun working with synths. My first keyboard and my first love was actually the Hammond Organ in our living room, and touch sensitivity for me has always been a weird thing to master. Which gets us into historical perspectives…..
The first synthesizer, in general terms, was the pipe organ. Its various manuals and stops allowed the tonal characteristics to be "programmed" by its user, to hint at non-organish sounds to some extent. Arguably, then, the first electric synthesizer would be the first electric organ, right? (Well, no, not right, actually. The first electric synth was the Telharmonium, a monstrous device that filled an entire building, ran its music over Con Ed telephone lines to "Telharmonium Parlors" in New York, generated a vast amount of relay noise and waste heat, and eventually burned down. Alas.) Anyway, the invention and proliferation of the electric organ produced one seemingly inescapable chain of facts, as follows:
- Notes must be activated by the player, somehow.
- In an electric device, you activate things by closing switches.
- Ergo, the player must play switches, one for each note.
This is the Switch Proposition. The second chain of facts states:
- The easiest way to design switches in an understandable array is
to use a keyboard. After all, that's how organ manufacturers did
it for centuries. - Ergo, it makes good sense, both from a standpoint of constructive ease and from one of familiarity, to put a traditional organ keyboard on a synth.
This is the Keyboard Proposition. Both of these modes of thinking are fallacious and full of holes, but they established the industry as we know it today rather than allow synthesis to go the way of the kazoo as a fad, so I guess we shouldn't complain. |→
The manufacturers couldn't have foreseen the trend they were starting; while many individuals utilized non-keyboard means of doing things (to be covered elsewhere), a large number relied on the cheap and easily wired organ keyboards for their research. A lot of good ideas that violated one or the other of the above propositions ended up enjoying little or no popularity due to the essential "alienness" of their design. It was nothing short of a miracle that the Ondes Martenot enjoyed the popularity it had at the early part of the century (Ravel, I believe, wrote a number of pieces for it), in light of the fact that it had a keyboard, but that the keys didn't move when touched.
Some designers relegated themselves to obscurity by their refusal to give in to the pressure of traditionalism: Lev Theremin, Serge Tcheprenin, Don Buchla, Louis and Bebe Barron, and so on. Others gave in to the pressure of keyboard use, but reserved a number of other means of controlling timbre, resulting in the fusion of keyboard and "Left Hand" control we see today: Robert Moog, Allan Pearlman, and Hugh LeCaine, whose left-hand controls for the Electronic Sackbut (It's a medieval trumpet of sorts, people. Get yer minds outa the gutters! |→ ) set a standard for expressiveness in real time that no one, and I mean NO ONE, has yet to equal. But the keyboard, as a means of selecting event start and note pitch, stuck, and gradually evolved into the form we see today. The pressures to produce an instrument that could be triggered reliably as well provide a pitch reference recognizable by some large percentage of ordinary musicians resulted in the vast majority of synths being controlled by keyboards. (The fact that the piano is a staple instrument in most music schools must have helped a bit too, as schools were where the first synths were used heavily.)
So where are we now? Well, the "average" synth keyboard has 61 unweighted plastic keys, each with a contact switch and some means (electromagnetic or optical) of sensing how quickly the key is struck. Some also can tell how quickly the key is released, and many have sensors under the keybed to detect pressure after the notes have been struck and held, either one for the entire keybed or tiny ones under each key. These are the familiar velocity, release velocity, and aftertouch sensitivities we demand from today's synths. It should be noted that they developed at differing rates; Moog made at least one synth that had pressure sensitivity without velocity sensitivity, and the Sequential Prophet T8, one of the first MIDI synths, had poly pressure rather than mono. When it went out of production, poly pressure languished as an unused part of the MIDI spec for years, until the Kurzweil and Ensoniq instruments revived it. There have been other modifications to keyboard design as well: see this month's issue of KEYBOARD for a bizarre but intriguing Japanese keyboard design using multiple rows of keys, all white, and consider the Yamaha GX1, whose keys bent pitch if you shifted them from side to side!
And, of course, the recent renaissance of piano technique in pop music has strongly encouraged the development of weighted actions of 73, 76, and 88 keys on many synths. There are even a few synths that have gone beyond this, such as the 95-key Bohm controller or the 97-key Beilfuss.
The keyboard has a certain monopoly over our electronic thought processes even today, though alternatives abound, but it's no longer either stagnant or monopolistic. In fact, I rather think I'd miss having one around: getting crazy is more fun if you have a reference point to ignore! |→
That's it for part 1. In future parts of this series, if demand indicates, we'll consider other means of applying traditional technique to electronics through special controllers, then delve into my favorite topic: realtime timbral control, via left-hand and other controllers. Look for the Electronic Sackbut, the Theremin, the Ondes Martenot, and the GX1 to reappear, as well as the Starwind, the Cyclone, the KAT, the WX7 and EWI, the Simmons SDSV, the ElectroHarmonix Drums, Moog's system 55, the Buchla synths, the VCS3, the EPS and VFX by Ensoniq, the Yamaha MCS2, the Mellotron, the IVL Pitchrider and Fairlight Voicetracker, the Grid, the Roland GR500 and its many children, the SynthAxe, the Oncor PolyTouch, and various ribbons, wheels, levers, buttons, sliders, switches, joysticks, paddles, mice, pedals, and hooters. ("Hooters"?)
Sounds good? Well, let me know if you're interested! I'll also be glad to elaborate on any particular instrument, answer any questions, or throw queries on stuff I don't know much about (there must be something here I'm not an expert on |→ ) to the List.
More, you say? MORE, you say? At ONCE, I say!
THE MAN-MACHINE INTERFACE, PART 2 PERCUSSION CONTROLLERS
Ah, drums. Whackety whackety whack! How we all love to hit things and hear them make noise… it's a primal instinct we all share. The love of percussion is as old as civilization, and it's a safe bet that the sorts of percussion devices we consider commonplace these days were elegantly refined millenia before the keyboard came along. So it was inevitable that when someone discovered that you could press a key on a synth and it would go "CRASH!", they might have begun to muse that it might be eminently (pardon the pun. actually, pat yourself on the back if you even spotted it! |→ ) more satisfactory to give some undefined THING a good whack and hear it go "CRASH!" instead. And the rest is history….
The first well-known electronic drum was the Moog percussion controller, a pad with an impact-sensitive resistance in it that could send a trigger signal and a voltage based on impact velocity to a monophonic synth. Each synth had to be patched individually for each pad, but the sound was (to audiences in 1974) never before heard. Check out Toccata,
from ELP's BRAIN SALAD SURGERY, or the live version on WELCOME BACK MY
FRIENDS…. for a real taste of history.
The late 1970's saw a huge proliferation of affordable analog drum pads, each with its own unique sound and character. There were the Synares (I almost bought a rusty old Synare 1 recently; they're rare as hens' teeth and sound great!) from 1 to 4, ranging from 4-pad synths to single drums, the Syndrum and Syndrum CN, the Pearl synthetic drums favored by Klaus Schulze, and (sigh) the ElectroHarmonix drums, a series of cheap, ugly devices that were a hell of a lot of fun to work with: the Space Drum, Clap Track, Crash Pad, Sonic Boomer, Rolling Thunder, and Clockworks "Percussion Brain." The Space Drum, with its awful descending "Pooooo….", was a huge part of what made disco forgettable, and makes it impossible to listen to the live version of Jean-Michel Jarre's Magnetic Fields II from LES CONCERTS EN CHINE without laughing fit to die. But drummers perservered; there was something here that they liked, and there had to be a way to get to it…!
And then along came the rhythm machine, and the music world began talking about the demise of the live drummer. Remember those days? Boy, I do. And I remember a guy whom everyone laughed at and made fun of, too. His name was Dave Simmons, and he had one hell of a good idea up his sleeve.
Trivia question: What's the first album in history to feature the classic Simmons sound? Give up? It's 1980's FROM THE TEA ROOMS OF MARS TO THE HELL HOLES OF URANUS by Landscape. Great disc, actually, and the Simmons drums were all over it. They really hit the scene when Bill Bruford took them on tour with King Crimson the following year, and since then, Simmons has defined the state of the art in electronic drums, from the old hard plastic ones to the zone-sensitive ones for the SDX…. and now they're out of business. Alas!
Technology advances, thank goodness, and the technology of pads is no different. The old sensors have largely been replaced with transducers and film-pack force-sensitive resistors, making the newer pads more reliable and tough than their parents. Also, the epidemic of "Simmons wrist" has led to more ergonomic designs for playing surfaces. As for expressive control, the real key lies in what sort of interactive software comes with the pads, for setting up gate times, response curves, and so on. After all, a hit is a hit is a hit if that's the only data you're feeding your synths. We've come a long way from the rubber discs of the Synsonics kits (remember them?) or the genuine leather pads of the MXR Kit (I almost bought one, in 1982… sorry I didn't, now…)!
Other notables in the "stuff to hit" genre included the Octapad and its many, many imitators (the original Roland design still providing the best pad-to-pad isolation in the business), the Kat and Silion Mallet for marimba types, and various bizarre MIDI devices like the Dynacord Rhythm Stick. (What a screech THAT thing was!) And although Simmons are no more, the modularity of MIDI has spawned a whole slew of percussion controllers accessing drum box sounds, samplers, and even (gasp!) the occasional analog synth. These days, the stores are again filled with cheap drum pads, but now they're digitally sampled and often have rhythm sequences built in ! But I'll admit that my tastes run to hitting an Octapad and listening to my Xpander go "CRASH!", but then again I've always been hidebound! |→
I think that a fundamental understanding of rhythm is vital to a musician's sensibilities, and there's something wonderful and visceral about actually playing a drum. That's why I find that the most convincing drum tracks on electronic albums these days tend to be done live, at least with an Octapad.
And even if you can't afford the cheapest MIDI controllers, like the $100 Yamaha DD5, you can always get a HandClapper or SpaceDrum and go wild!
Next, in Part 3, we get a little more serious, as we discuss more serious technical problems in a look at stringed instrument control of synthesis. Here we go again, folx. Feedback on Pt 1 was largely positive, and Pt 2 didn't seem to generate much comment at all, so let's see how we do this time…
* THE MAN-MACHINE INTERFACE, PART 3 STRINGED INSTRUMENTS
The guitar and its cousins are, in comparison to keyboards and drums, huge pains in the anatomy to use as control devices. Why? Because they violate both the Switch Proposition (each note is not an event in itself) and the Keyboard Proposition (There's no logical layout for pitch reference in a mechanical way). But guitarists are an insistent lot, so let's look at the history books, but not until after we have a quick tutorial on the technology!
There are two recognized means of controlling synthesizers from a stringed instrument. One is to mechanically analogue the control surfaces of the instrument and map them to a synth control interface, and the other is to read the actual notes as they are played and interpret them for the synth's use. These two procedures, which (for reasons that'll be obvious in a moment) I'll call "fretwiring" and "pitchtracking", each have strong advantages and disadvantages, which should be enumerated here.
FRETWIRING has the strong advantage of being as quick and responsive as the keyboard on a traditional synth. The analogue reads your actions as fast as you can act, and the synth responds at the speed of light. In addition, it is easier to directly incorporate control of the timbre or nuance of the synth tone into such a design, merely by adding the appropriate sensors for the types of mechanical movement the player will produce as he plays. In other words, the instrument follows one's movements perfectly and instantly. The disadvantages of fretwiring are in its incredible complexity and expense, as the guitar neck must be wired to sense fret contact by each string (hence the name) and other performance sensors must be able to quickly and accurately follow the guitarist's movements. Also, it is rare that such an instrument can function properly and yet still serve as a real electric guitar, making a synth/guitar blended sound impossible. Lastly, while the possibility of direct timbral control of the synth exists, it must usually be implemented at least in part by controls that are not common parts of the guitarist's technique.
PITCHTRACKING relies on circuitry and special pickups that follow string movement and determine the pitch and amplitude of a note being played (hence the name). The string is plucked, the pickup or other sensor reads its relevant data, and the synth is informed accordingly. The advantages of this system are its ability to be easily and nondestructively mounted to any good guitar, allowing for a natural playing style and a blend of guitar and synth timbres, and its relatively low expense and ease of upkeep. Its primary disadvantage is in the tracking process itself: it takes some finite amount of time to read a guitar's pitch, and more time to translate that pitch to a voltage, and more time to go from voltage to MIDI. Thus, guitarsynths can be noticeably slower and sloppier in the synth output than in the straight guitar sound. This problem is compounded by the fact that many guitar nuances serve only to confuse the pickup, imposing an artificial and stilted playing style on the player. Lastly, any performance controls that the synth alone will require must somehow be attached to the already cramped guitar body, causing potential confusion and clutter.
So how did these two approaches fare through history? Well, the early years were dominated by pitchtrackers, as the pitch-to-voltage converter was a neat technical problem that folks had been working on for years. Early synths, such as the Korg X911, were monophonic: only one cleanly-played note could be played at a time. The ARP Avatar was another example of this strategy. Sadly, these instruments were ahead of what the technology could do, and while the Korg was only an embarrassment, the Avatar contributed strongly to the death of ARP, alas. The first Zeta pickups appeared at about this time, for use with outboard P/V converters and synths. The late 1970s and early 1980s saw a profusion of these gizmos, and few if any worked well. Notable, though, were boxes like the ElectroHarmonix MicroSynthesizer, which was nothing more than a square-wave divide-down box, and the Oncor Touch guitar, which was the first fretwired instrument in production: it had metal bars in place of strings, and was actually an oddly-shaped monophonic synthesizer!
The guitarsynth's evolution progressed slowly into the 1980s, and might still be progressing today, if not for the interest taken by one particular company in the challenge: Roland. Their first attempt, the GR500, utilized a bass synth, a mono synth, a string machine, and a divide-down polyensemble, all controlled by a hexaphonic pickup on a very nice guitar. It tracked half decently, it sounded great, and it was a big hit at once. (The guitar parts on "Follow You Follow Me" by Genesis on 1978's AND THEN THERE WERE THREE are all GR500.) Roland followed this up with the GR300 series, which Robert Fripp made his own (see any King Crimson album from 1981 or after), and eventually, the GR700, which was a breakthrough in three areas: it was a true polyphonic guitar synthesizer, it could be set up reliably and personalized easily to each guitarist's touch, and IT HAD MIDI! Since then, Roland's main drive has ben to get faster and faster tracking pickups, and to an extent they've succeeded. Ibanez has also marketed a MIDI guitar, with a MIDI whammybar!
The fretwired guitars were much later in coming, but they had advantages. Muting didn't faze them, nor did hammerons and pulloffs, and unlike pitch sensors, if the neck was designed right they didn't mind bends at all. The first and most famous of these was of course the SynthAxe, an $11,000 MIDI monstrosity that was enormously powerful and expressive, but actually required a fair amount of retraining to use properly. (Allan Holdsworth's ATAVACHRON and SAND are loaded with SynthAxe work.) The Stepp DG1, though the company died, was an improvement on the SynthAxe in terms of familiarity of use. Other designs that attempted to avoid pitchtracking problems utilized sonar or light sensors (the Yamaha GS1 (?) and Beetle Quantar), but these combined the disadvantages of limited fretwire control and no real guitar sound and hence didn't catch on terribly well. On the flip side, Zeta is marketing a guitar now called the Mirror 6, which is a real guitar that at great expense has been totally fretwired, giving the ultimate in both sides of the coin. The instrument's cost has been offputting, but field reports say it may well be the wave of the future.
As for other instruments in the string family, the bass guitar has been approached only by Roland and Zeta. Roland's attempts to quickly trck the low-frequency bass notes have been frustrating, but Zeta has hinted at a Mirror 4 bass to be released. Zeta have also developed a violin synth, utilizing a special pickup. Since there are no frets and the bow doesn't always start a note cleanly, this instrument uses pitchtracking, and does fair to middling in following synths. Its real strengths, according to users like Jean-Luc Ponty and Emilysue Pinnell, are as an electric violin, which can be transposed down into the cello range without difficulty.
If I've missed any designs, feel free to remind me of them. But as for the future, my money's on a dichotomy of designs: some folks only want a guitar sound with something else behind it, and for them the Roland designs will do fine for cheap, but others want the precise control of fretwiring and will pay to get it. That being the case, I think that real-guitar-based designs like the Zeta will eventually win the day, as they do leave open a chance to play the real thing. As prices drop, we'll see who comes out on top, if any one type does. In the meantime, we'll have to get used to guitar runs with string chords tagging along behind them, I guess….
Part 4 will be delayed for a trip to do research elsewhere in the USA. But when I get back, we'll look at wind instruments and the problems they pose for synthesis. I'll also be taking suggestions for other instruments to cover, because if I don't get any, then in the fifth and final part I'm planning to cover all of the bizarro forms of synth control that have come along over the years, from the pitch wheel to the HANDS. Well, here we are in Part 4 of my little series. Feedback has been sporadic but generally friendly, so I'll quit while I'm ahead (after this installment and the next one, which will cover unusual timbre and nuance controllers).
* THE MAN-MACHINE INTERFACE, PART 4 WIND CONTROLLERS
The wind and brass instruments fall in between stringed and keyboard instruments in terms of translational difficulty: while there are sets of keys that definitely tell the instrument what note is being played, the nuances of breath control are perhaps more complex than those of string playing. Lip and tongue pressure, expulsion of air, and jaw stiffness can control vibrato, loudness, tremolo, even pitch bends and octave jumps. But the fascination of combining this sort of expressiveness with the sounds of the synth has fascinated people for decades, and a number of attempts were made:
The two most famous historical wind controllers were the Lyricon and the EWI and EVI. The Lyricon, marketed by Computone, was a clarinet-like controller that allowed pitchy and vibrato control over analog synths through variable control voltage levels. The EWI was a similar development using a different approach to control style, and the EVI was a design for trumpeters. While to hear the EWI you'd need to dig up albums featuring people like Michael Brecker, the Lyricon was actually a fairly well-known beastie: Emerald Web used them a lot, and the Tangerine Dream album CYCLONE features one. Another device was the selfcontained Starwind, a mono synth with a melodica keyboard and elementary breath control. Neat device, and inexpensive, but never caught on. It would take the advent of MIDI to produce the same sort of revolution in wind control that had occurred elsewhere in the industry.
The primary advantage of these early devices was that they utilized analog signal control, hence had infinite degrees of control shading. Digital control, by definition, was coarser, and the Yamaha WX7, the first MIDI wind driver, suffered from some complaints in this regard, as does the newer, simpler WX11. The newer EWI and EVI, now made by Akai, actually run a special synth via analog voltages, hence preserving the expressiveness. They, however, suffer from a non-intuitive keying system that many players find unpleasant to use. Other instruments include the Synthophone, the Atrisyn MIDIsax, and the Sting. Each has its own approach to handling the dual needs of familiarity to seasoned players and of faithful synth control, and each involves tradeoffs in terms of performance control, expressiveness, and cost. The WX11, for instance, has a far more limited set of expressive capabilities than any other wind driver as of now, but it's simple to learn and quite cheap. The Synthophone, on the other hand, is built inside a real sax, but suffers
from high cost and limited synth control. The WX7 is considered
difficult to "set up" by individual players, and the MIDIsax doesn't even allow certain types of setups to be performed, though it does have a number of useful features like expression controllers and a program readout where the player can see it easily. And the list goes on.
The ultimate test, though, appears to be in the hands of the player. While the guitar synths have divided their users into two clear camps, the wind drivers are so diverse in their capabilities and limitations that it seems neccessary for the individual to do a great deal of legwork before making a personally satisfactory choice. These days, the two best sellers are the EWI (due to its expressive synth control and Akai's strong commitment to making it succeed) and the WX7 (customizable to each player's embouchure and touch, universal in MIDI control use, and now quite cheap now that it's out of production). We'll have to see what the future holds. Oh, and what synths are normally used with them? Hmm, the Akai comes with its own, and the WX11 does too (It's suggested to use a TX81Z with the WX7), but outside of those obvious choices, the strong favorite is the analog-controllable, MIDI-compatible, UNBELIEVABLE-sounding Oberheim Xpander. It's also got a strong following among guitar synthesists in both camps, for similar reasons, although I know of no one using it heavily for percussion applications. Its rich analog sound and fine controllability by multitimbral devices make it a natural. (Forgive me, folx; I have two. |→ )
Next, our fifth and final part will cover timbral controllers of various sorts. If you have a bizarre favorite, please let me know, and don't be shy about which category they fall under: I'll be covering oddball keyboards, pads, stringed devices, and wind drivers as well as other stuff! Well. Here we are at the end of this little series. If there's anything I've missed, please feel free to ask me, and I'll include it later. But for now….
* THE MAN-MACHINE INTERFACE, PART 5 UNUSUAL CONTROLLERS
When I say "unusual" I mean stuff that either puts a new wrinkle on old ways of determining pitch and start/stop times of events, or entirely new methods of doing so. I'll be skipping all over the place, and I'm sorry for that, but it was either unformatted or never to be finished, so what the hey.
The pre-MIDI era, and in fact the pre-digital era, abounded with bizarre attempts at control devices. As has been mentioned elsewhere, the Ondes Martenot had a ribbon attached to a ring on the player's finger, that slid back and forth as he played and served as a control device in that fashion. But for the most part, the pre-vacuum tube age relied mainly on old standbys of the organ design palette for control: stops and keys.
With the development of vacuum-tube synths, and then solid-state devices for electronic tone generation, came a new generation of devices for realtime shaping of tone and timbre. The simple organ-stop designs of early keyboards were cast aside in favor of a plethora of new approaches, the traditional keyboard was cast aside with somewhat less success, and in some cases both were simultaneously cast aside in favor of a new, combined tone-timbre control system. For example, I'm sure that everyone here has at least heard of the Theremin: it was a marvelously expressive instrument in the hands of a skilled player, and they were turning up in various and sundry places for decades after their invention. The Theremin consisted of two antennas, one horizontal and one vertical (at least in traditional form). The player held one hand over the horizontal antenna, and waved the other hand near the vertical antenna. The horizontal antenna controlled loudness and the vertical one controlled pitch, based on the proximity of the player's hands to them. As late as the early 1980s, there were Theremins available to generate CVs for synthesizers, and there exists a Theremin society even today. Jimmy Page uses one in THE SONG REMAINS THE SAME, and at least one version of the theme from STAR TREK uses one as a solo instrument. I had a chance to play one a few years ago; it's not as easy as it sounds, but the degree of nuance and subtlety were obvious even to a neophyte like myself.
Other attempts to remove the early synthesizers from their dependence on traditional keyboards took the form of, well, alternate keyboards. There's been an ongoing argument as to whether or not the ole black'n'whites are the best means of controlling ANY instrument of a keyboardish nature, including the piano and harpsichord, and a number of acoustic pianos and synths have been developed with alternative keyboard designs. The Vandervoort keyboard uses a design similar to that developed by von Janko earlier in this century; it uses four rows of keys in whole-tone series, allowing for consistent fingerings and easy multi-octave jumps regardless of key signature. It was designed in 1976 and was originally envisioned as a "shell" to be placed over a traditional keyboard; what's funny is that KEYBOARD wrote of it as a failed and forever lost idea in their Feb. 1987 issue (A great reference, by the way), and just last week, MUSIC TECHNOLOGY mentioned the Vandervoort in a working prototype form as being on display at last summer's NAMM show– complete with a photograph! Ironic, isn't it? Another, less well-known design was the Secor keyboard, which utilized a honeycomb of 240 multicolored hexagonal keys to allow easy (sic) use of microtonal scales in performance. Its inventor, George Secor, said of it that for less than 31 tones per octave it wasn't worth learning, and that even if it never caught on it would look great as a prop in a SF movie. That's for sure: only two Secor keyboards were ever built, for a little-known synth called the Motorola Scalatron, and they were BIZARRE! Beautiful, but bizarre.
Another proponent of alternate designs was Don Buchla, whose wonderful synths almost never featured traditional keys. (One exception was the Touche, his one attempt at a commercial machine: more on it in a moment) He instead preferred capacitance touchplates, which were pressure sensitive yet had no moving parts to break down. A Buchla synth was a sight to see, with arrays of keys, controllers and expression devices laid out as no more than patterns of wire on a flat plate: almost like playing a bas-relief sculpture of a keyboard. The design was simple, reliable, and infinitely flexible, and he still uses it heavily today. Its one drawback to most users was its lack of tactile feedback; it didn't move when you touched it.
Nevertheless, the touchplate design was inexpensive to build and reproduce, so if feedback was less important than economy, it got used– a LOT. Entire 1980s instruments in the pre-MIDI era used touchpad keyboards: the ElectroHarmonix Mini Synth (of which I proudly own one, and love to death), the EDP Wasp (Black and yellow keys…eegh!), and one oddity called the Dubreq Stylophone, which required that a stylus be used to touch each key. (It didn't catch on.)
For people who insisted on the old-fashioned piano keyboard, there were still a lot of frontiers to cross: one manufacturer has marketed a keyboard which not only has velocity and pressure sensitivity, but where each key can move in and out or side to side! Watching one of those in action is kind of frightening; like someone put LSD in your tip jar…. The Yamaha GX-1, the half-ton monstrosity made famous by Keith Emerson, had at least one keyboard manual that allowed side-to-side key motion. (As an aside: I'm looking for data on this beast: I know nothing of what it could and couldn't do, how many were made, or how it worked. All I know is, it weighed nearly 1000 pounds, had two fullsize key manuals and one minikey manual, and Keith Emerson had two: one which got destroyed by a runaway farm tractor (no, really!) and one he got from John Paul Jones of Led Zeppelin. Stevie Wonder had one, and so did Jurgen Fritz of Triumvirat (who was famous for having one of everything Keith had, no matter what). Other than that, I'm stuck. Help!) And then, of course, there were always the feet: organ bass pedals have been in existence for centuries, and from the Moog Taurus to the MIDIstep they haven't been far from the electronic world either.
Beyond the realm of actually playing the keyboard, however, was the area of controlling the sound as one played. The area of timbral control is even richer than the area of keyboard triggering, and has spawned all sorts of approaches to seemingly trivial problems. Consider the two most common ways to alter a sound played on a synth (not including the sustain pedal, which is one thing that survived unscathed from the piano pedagogy): pitch bending, or the addition of vibrato or tremolo, i.e. LFO modulation of frequency or volume. "That's easy, man! You just get these two things on the left side of the board, and…." Yes? And do what? Pull them, push them, wiggle them, stroke them, press them, punch them, squeeze them, what? And in which direction? Up, down, left, right, in, out, forwards, backwards? And what do they do when you let go: return to zero? Stay put? Keep going to some preset destination without your help? Hmmm? Well, the answer to all of the above questions is, of course, "Yes," so it might pay to look a bit at the simplest (HAH!) of control surfaces before we get on to the heavier stuff….
The proliferation of left-hand controllers on the market, today just as badly as yesterday, is due to three factors: ergonomics ("What feels best to a player?"), functionality ("What'll work best for a technique of play?") and, alas, marketing ("To hell with Items One and Two: What'll make us look different than everybody else?"). In the beginning, pitch and expression control was performed by varying control voltages on modular synths, and this has persisted to the later, hard-wired synths. Nowadays, it's done with digital data messages, allowing for some new wrinkles, but the fundamental ideas haven't changed much. You're holding a note, and you want the timbre to DO something: so what do you do? Well, on the old modular synths, you grabbed a knob and turned it. Simple enough, but it lacked something. So the earliest synths for the mass market, the Moog modulars, had a ribbon controller: you put your finger on it, and the pitch jumped up a bit. The jump depended on where you touched it, and when you let go it jumped back. Simple, right? Well, there were a lot of lead synthesists whose techniques relied on them, and when Moog stopped putting them on their machines, Yamaha did: the only MIDI ribbon available today is on their KX5 strapon. I love it!
The replacement for the ribbon, of course, was the wheel, and this is still (in its MANY forms) the most popular type of controller today. How do they work? Well, that depends on the synth. For an old Moog, the left one's bidirectional and is used for pitch bends, with or without a center detent, and the right one's unidirectional and is used for LFO modulation. This "standard" (with the modern additions of a spring return-to-zero on the pitch wheel) is used by Ensoniq, Emu, Yamaha, and on older Roland and Korg boards. Oberheim uses a reversed design that was copied by Chroma and (partially) by Kurzweil: the wheels are shaped like paddles, and both are bidirectional and springloaded. Older designs for such a mod wheel were "dead" in one direction; nowadays, each direction accesses a different modulation (VCO vs. VCF, perhaps). Roland has a side-to-side pitch paddle, which is pushed forward for modulation: I, personally, find this dsign unusable and not a little silly. Korg uses a joystick for pitch bend and two types of modulation (left, flat; right, sharp; up, VCO; down, VCF) that wouldn't be bad if the joystick moved in a circular well rather than a square one and wasn't so flimsy, argh! But I digress. The joystick idea was also marketed as a pressure plate, where the XY position of the fingertip sent out voltages. Another idea that ARP tried was the PPC, or Proportional Pressure Controller: three pads, on which you rested your left index, middle, and ring fingers as you played. Pressing one bent up, pressing another bent flat, and the one in the middle added modulation. (It didn't sell.) Then there was Yamaha's old setting-the-clock pitchbender, and… You get the idea. One other expression controller, that's been single handedly shoved into the spotlight by Yamaha, is the hooter: the Breath Controller. It's a gizmo that generates a CV based on breath pressure, allowing lines to be articulated as with a wind or brass instrument. As Chick Corea once wrote: "You get into blowing." The Yamaha MCS2 specializes, in fact, in adding controllers to MIDI pianos and the like: you get a MIDI merger that adds a pitch wheel, mod wheel, hooter, two footpedals, two footswitches, two sliders, and three buttons to your MIDI control arsenal. Yippee!
Of course, this doesn't hold a candle to the single best design for left-handed control of a synthesizer ever, which was invented by Hugh LeCaine in the 1950s for his Electronic Sackbut. The left hand rested on a plate, with each of the four fingers controlling a slider that varied some component of the sound in realtime, so flexing the fingers caused a very smooth and natural evolution of acoustic-imitative sounds such as strings. And the thumb, not to be outdone, rested on a sliding pad that moved joystick-like across a circular plate that was divided into different waveform types, and allowed formants and waves to be blended in real time with a twitch of the thumb. Only LeCaine himself ever mastered this instrument, but the sound was, well, amazing. The imitation of a string quartet he did in 1958 was frighteningly real, not so much in sound but in articulation of the instruments' bowstrokes!
Other parts of the body were used for timbral control, of course: footpedals and footswitches were and are common, and the GX-1 had a gizmo for bending pitch when the player drove his knee up under the keybed. Visceral!
Now, one off-the-wall area of timbral control that deserves a mention, simply because it's an idea that just won't die (for obvious reasons), is that of event triggering as part of the performance controls. For example, instead of pressing the pitch wheel up a half step and hoping one doesn't overshoot, one presses a button that tells the synth, "Bend up a half step, at such-and-such a speed." This idea could be applied to anything: sequences of notes, chords, vibrato or other modulatory effects, and so on. And although the obvious way to do such things these days would be with a MIDI sequencer, the idea's far older than that. An early strap-on synth called the Syntar (used a lot by Michael Garrison, among others) had nine such buttons on its "neck" as user- settable triggers, and the Buchla Touche was designed to operate in this fashion
from the ground up. Nowadays, we have the Roland Axis, whose thumb
button or finger buttons or footswitch can be set to slowly bring in Controller 1, and the MIDI Mitigator, where every footswitch can deliver up to 256 bytes of user-choosable MIDI data… even System Exclusive, so that you can turn a knob on your– wait a minute! Isn't this where we came in? |→
And then, almost exclusively thanks to MIDI, we have the explosion of percussion controllers; it's now possible to create drums sounds with nuance and timbral variety from a number of sources. The Acoustic drummer can supplement his kit with an electronic "box-o-pads" like an Octapad, DrumKat, or Portakit: not a new idea, as MXR was marketing an analog version nearly twenty years ago! And there's the Rhythm Stick, a strapon drum controller you fret to select sounds and hit to play them, and the PKI GunDrums, which are fired like pistols to play sounds selected by the thumbs. Triggers are built into the heels of shoes, into glasses frames, into body suits: there's no limit now to the event control of MIDI via impact.
And what of the future? Well, we're seeing bits of it today, in the Airdrums, a pair of MIDI batons that send messages when moved on their axes in various directions, and the Hands of Michael Waisvicz that control synth parameters in real time by flexing the fingers, touching the palms, waving the arms and turning the hands (I want a pair. BADLY!). Robert Moog recently spoke of a design student's idea for a pole-shaped controller with keys on its grips, played in a dancelike motion. And for the instruments themselves? I have fond memories of the instrument that sank PPG: the Realizer, where you punched a button and the entire front panel changed. Bip! It's a Minimoog. Bip! It's a DX7. Bip! It's an Emulator….
Yes, this is a good time to be an experimental musician. Maybe most of you will never take the time to wonder about all of this "man-machine interface" stuff, but I know that my creativity goes way up whenever I do… It's just part of the process with me. And I hope that maybe I've given some of you some food for thought. Thanks for listening.
Oh, before I forget: I did get one contribution, from Pete Lucas:
* Man-machine-interface: One instrument i came across some years back was a thing called a 'Tactophone'. It was a metal framework (steel tubes and angle sections bolted together, standing about 3 feet high) with wires stretched between the upper and lower members, in the style of a harp. It had about thirty strings if my memory is correct. An electronically energised coil vibrated the top member, and hence the strings, there were pickups on the individual strings along the bottom edge. Output from the pickups was fed back to the energiser, hence if you plucked a string, and then took your hands off the instrument, you could get a sustained note, or a note that died away very slowly, depending on the settings of the amplifiers (oh yes, i forgot to mention, there was a mass of amplifiers, ring modulators, notchfilters etc attached, with a console for the sound engineer to manipulate) You could run your fingers up and down a string, instead of plucking it, this produced noises like a musical saw (ethereal wailings reminiscent of the ondes martenot, or running a wet finger round the rim of a glass), or alternatively, since some of the strings were wound, running a hard metal object up & down the strings made interesting grating/screeching sounds. If you got the gain of the amplifiers just right, then merely touching a string very lightly would start off a vibration, which would grow louder and louder and louder. Brushing a feather across all the strings sounded nice. There was a pedal attached to the electronics, which cut the amplifier (and i think also introduced negative feedback, coz it seemed to damp the vibrations very fast indeed). I also saw the thing tipped on its side (strings horizontal) and played like a xylophone with two metal rods as 'sticks'. The 'player' and 'composer' of music for the beast was, if i remember, called David Morkrum (or was it Morkum, Morecambe?). Wonder if you've heard of this 'thing'? Possibly under a different name?
Well, thanks for your attention. I hope folks got something worthwhile out of all this; I had a lot of fun going back through my archives getting this stuff out. Now it's time for me to go back into the studio; I just had this great idea for a song, using nothing but a pitch bender….