THE LEARNABLE INSTRUMENT PROJECT
copyright 2001 Chris Wayan
To design musical instruments simpler to learn than traditional ones, and yet able to play a full repertoire of tones, scales, chords, and types of sound. At present I'm exploring nine general designs:
Acoustic instruments must do two jobs at once--produce good sound, and have reachable and usable controls. The first job often severely constrains the shape of the instrument, and the variety of sounds is limited. Since electronics entered music, a lot of work has been done on the first half of the instrument problem: producing a wide variety of sounds and shrinking the bulk of such clumsy monsters as the piano. But the second half of the problem has been neglected: the interface. What is the most logical, learnable control system? The question's rarely even posed; most innovations have remained novelties. But the potential is there: electronics allows us for the first time to free the controls from the constraints of a sound generator, so the input device can be shaped to suit musicians not acoustics. We need better physical ergonomics (ease of reach and handling), but especially MENTAL ergonomics (consistent, easily learnable patterns).
The challenge is to find something fundamentally different from the three commonest designs inherited from classical instruments: a piano-style keyboard, a neck (fretted or not) with several strings to be fingered, and the sets of holes or valves found in wind instruments. These range from merely awkward, to complex, arbitrary and physically exhausting, to downright painful (ask any beginning guitarist). We can do better: the room for improvement is vast.
Before the invention of the keyboard, the balance between blown, struck and plucked instruments was substantially different. Dance music could be quite loud, but most of the volume was from drumming, fiddling (which can be quite piercing) and horns or nasal woodwinds. No single instrument dominated or was all-purpose; the nearest thing to a universal or "default" instrument was the human voice.
Once keyboards became popular, they became such a default instrument--more precise than a human voice, with a wider range, capable of chords or polytonality, and with all possible notes spread out and visible at once: ideal for composition! With all these advantages, the expense and bulk and tuning problems and slow learning curve were a price worth paying. The piano became the musician's and composer's workhorse for over a century.
In 1945 no one would have ranked the guitar as anywhere near as important as the piano. But with amplification even a single guitarist playing simple chords could fill a room. The advantages of the new sound were several: beyond mere volume, it could be altered in far more ways than a piano's mechanical pedals allowed: bending and reverb and flanging offered immediate sensual delights. The sheer variety of tones could only be matched by a good singer. While full mastery was slow, one could learn a few chord positions and begin playing at once, acquiring others gradually. The disadvantages, such as the physical awkwardness and arbitrary quality of some chord fingerings, and the need for an amp and a power source, were acceptable to millions. Guitars dominate most working bands today, and almost all pop music scores mark the guitar chords as well as the piano part--if they bother with piano at all. Yet this huge shift in popularity was driven by just one minor innovation: amplified output. Though guitar-builders quickly realized that resonance chambers need not be large, so electric guitars could be slimmed down radically, the basic control remained the fretboard, barely changed for centuries. Some guitarists invented alternate tunings, but these are analogous to "software" innovations: the hardware, the basic structure, remains the same: cramped, awkward (even painful) to play, and varying in size on different instruments.
Pianos, needing no amplification, changed very little in midcentury. Even when synthesizers shrank to a commercially viable size and started to replace acoustic pianos, the main push was to offer a variety of sounds and effects. Acoustic pianos had never bent notes or had reverb--these are direct borrowings from the electric guitar, which had a twenty-year technological lead. Hungry to catch up, keyboard makers scrambled to enrich their output until synth sounds were more varied than any electric guitar--indeed, that's their main selling point today. Learn keyboard and you can play a thousand (cheesy) instruments (badly). But just like the guitar, its input section hasn't changed since 1700. And why should it? The musicians likeliest to buy new, high-end products are usually skilled at the current interface, cumbersome though it is for beginners.
The closest analogy, and it's an embarrassing one, is the standard typewriter keyboard--built intentionally awkward, to slow human typists down to the pace the links and keys could handle without colliding. Mechanical links that are no longer there! But once a generation of typists had learned one pattern, it's hard to switch to another. The Dvorak keyboard really is more efficient--I write as one who's tried it. But few use it--because then you MUST use a Dvorak keyboard; it's difficult to jump back and forth. The result of this legacy from the industrial age is built-in inefficiency a century later--tolerable in typing, perhaps, but a real problem in music where time and fluency is of the essence. In the last decade, ergonomic computer keyboards are at last appearing: the one I'm typing this on is reprogrammable to either standard or Dvorak (or Cyrillic for that matter). It's bent in the middle; keyboards with humps are increasingly common, and separate hand-pads are no longer experimental. As machines get more flexible, innovative designs don't lock you into a hardware ghetto, as they once did.
So I've been considering input devices in general, trying out patterns free of the constraints imposed by sound-generators, such as size, shape, or order of valves/keys/controls. An electronic instrument, unlike any other, can have controls built entirely for the convenience of the musician.
SOME POSSIBILITIES ----------------------------------------------------
DEVICE 1: THE SMART MIC
Picture a microphone translating your voice into MIDI commands, so it could be used to drive any digital sound generator just as a keyboard can. Crude voice-transformers and filters do exist--remember Laurie Anderson turning her voice into a man's? This isn't an easy problem for software to solve fully--recognizing what's the fundamental and what's an overtone, deciding whether to clean up the pitches or translate all the shadings, distinguishing the "target" voice from room noise, especially if a band is playing... It's the whole speech recognition problem squared, and it MUST be solved in realtime with no major errors. In the long term this will be commercially and artistically viable (there are millions of klutzes who can't master hand-controlled instruments who CAN sing or whistle, who would LOVE to have their karaoke machine turn their voice into a bell or a foghorn or their favorite singer's voice) but it presents serious technical problems now. (The commercial ones I've tried couldn't understand whistling, for example). Plus this is a software/computation problem, so any pioneer faces instant copycats: once a realtime voice-parsing device proves profitable, rival companies will come up with analogous software pretty quickly.
Also, the voice has disadvantages as an inputting mechanism: unless you're Tuvan, it's hard to generate chords--or even widely separated notes. You can input more data with ten fingers than one voice. So if a fingered input device is easily learnable, most musicians will prefer it. Look what they endure now!
So for now, I'll focus on fingered devices, keeping in mind the mutant microphone lurking in our future.
FINGERED INSTRUMENTS ----------------------------------------------------
What we want is an arrangement of keys or buttons so simple that you play the way you sing--without thinking too much, without worrying what key you're in. Your fingers should simply move a distance up or down to strike a run or chord: the same fingering no matter where on the instrument you play.
How to arrange the control buttons? Instrument makers can learn from another industry. The history of data storage and access, from papyrus scrolls to computers, is full of recurring solutions to the war between two urgent needs: compactness and accessibility. These solutions are:
1--the straight line folded or sliced and stacked into grids and the grids stacked into blocks, as in books, musical scores, RAM, or Web pages. This solution offers easy access at the price of discontinuity. You hit the end of the line, turn the page, wait for the screen to load...
2--flat spirals like audio or video disks; information is both compact and continuous, but they have a scaling problem: center and rim areas may read differently. Do you compress the center so things line up in a radial grid, or let them skew?
3--wound continuous scrolls like audio and videocassettes, papyrus scrolls, and film stock. These are simple and continuous, but it can be awkward getting to things--the rewind problem. A variant of this is the helix, seen in stelae (ancient columns with spiral stories drawn or written on them), early Edison recordings (wax cylinders), and, most famously, DNA. These offer continuity, full access, and no scaling troubles. Their limitations: they're not as easily flattened, so they're bulky compared to blocks and disks. Plus, you can't see all sides at once and they're harder for the human mind to visualize than grids or lines. They look and feel, well, screwy.
A piano is a Type 1 access device, though a subtle one. It slices the chromatic scale into an eight-tone white scale and a pentatonic black scale above it. The staggering of black and white keys does two jobs: it shapes the player's choices toward familiar diatonic patterns AND compresses the overall scale a lot, allowing big comfortable keys. Without that compression, only giant hands could span an octave. But it comes at a heavy price--the massive learning problems created by the irregular slicing. Every scale, chord, or melody manifests differently when transposed to a new key. Learning takes years, just as arithmetic would if we still counted in Roman numerals or if we measured by pecks, yards, ounces, rods and tons. (Oops, if you're American, you probably still do. Congratulations. How many rods are there in a mile? How many yards? Quick, this is a performance! The rest of the world went metric for a reason. Well, to push the analogy... most existing instruments let you play the major C scale in inches at the cost of playing the C# scale in centimeters and D in picas.)
A guitar is also basically a Type 1 device. It slices things differently, and overcomes some problems by allowing several ways to play a note, using different strings. It runs into problems in that the easiest notes and chords are down near its bass end, so one constantly hits bottom ("open" strings), so you can't lower them further. Going from a C major chord to a C minor involves raising notes instead of lowering them! Perverse. Some keys and chords are much harder to finger than others, and the same basic chords (major, minor seventh, augmented, etc) have no consistent fingering patterns--D minor is fingered nothing like E minor, etc.
The only patterns of keys, frets, buttons or holes that don't prefer some key signatures or chords over others (forcing you to learn many fingerings for the same type of chord or scale), are those that avoid breaking up the line of keys/controls into segments at all. Again, I see only three general solutions:
1--a straight line, like the frets under a single guitar string, or a truly chromatic piano. Problems: without slicing you may end up with keys too small for your fingers... or a seven-foot keyboard. Yet careless slicing creates severe learning problems.
2--flat concentric spirals, like a nautilus shell--analogous to data disks. Conical designs, like sea-snail shells, are topologically identical, and worth considering.
3--helical coils, like a winding stair, or DNA, analogous to scrolls or reels of data.
DEVICE 2: THE LINEAR CHROMATIC KEYBOARD
This design has the engineering advantage of being structured much like a present piano--a keyboard could very cheaply be modified to have twelve keys per octave, each the width of a piano's black key (about 14 mm). But that's pretty narrow--it's hard to have keys wide enough to play cleanly and still let one hand reach a whole octave comfortably. Still, there are far more awkward instruments out there, and skilled pianists could even adapt to it, since all the intervals would be the familiar size--only the recessing of black notes and their altitude difference would disappear. While fingering would have to be precise with such narrow keys, it'd be less cramped than on guitar necks, and fingering patterns for scales or chords would shrink from twelve to one. Mastery of such a keyboard should come far more quickly for new students. Such a machine could shift (using software only) from a 12-tone scale to Indian scales with 16 tones, or microtone scales a la Harry Partch, for that matter. With no built-in periodicity as on the piano, the octave can stretch or shrink freely--it's just a row of generic input switches. Just as with singing, the instrument here doesn't impose a key preference--for good or bad.
No walls in your way, but no guidance either.
DEVICE 3: THE OFFSET TWIN KEYBOARD
ANY fingering or inversion for a chord or scale you learn on the Offset would be valid for any key.
It's counterintuitive, but the ideal amount to offset the two rows might be a tritone (6 semitones, exactly half an octave): then you can just as easily play the UPPER keyboard and use the lower one as the spare, getting the exact same "boost" downwards as upwards if you cross over. The two keyboards would thus not really be master/slave but two equal partners. However, an offset of a fourth or fifth would also work quite well. Fooling around with such designs, I've found I prefer a fifth, but that may be just me.
The offset twin board also allows tunings difficult on a piano such as a 16-tone scale. Or one could set the upper keyboard for quartertones between the lower board's notes for Indian and Arabic music.
A triple board, upper row a fifth or tritone above the center, lower row a fifth or tritone below, would be even better, though I don't know if the added cost would justify it. Jankó, the last inventor to tackle the problem commercially, DID have three full stacked chromatic key-sequences, and found it did help playing.
The only reason I haven't focused exclusively on chromatic flat keyboards like these, is that it's an obvious reform that doesn't require modern electronics. Many inventors have tried over the last two centuries, yet even the very logical Jankó design failed to catch on, even to the small extent that the Dvorak keyboard has among typists. Still, a well-designed chromatic keyboard can be marketed far more cheaply than the Jankó keyboard was, around 1900. A hundred years is a long time for no one to try... and I've never even seen a design with two or three offset chromatic rows.
DEVICE 4: THE PLUCKER, A SHAMELESS MARKETING PLOY
It might help if a chromatic design was marketed not as a better piano but a distinct instrument particularly well-suited for chords--a painless guitar. I've experimented by tweaking the software of a standard synthesizer so that a sound begins only when a key is RELEASED (like plucking a string). The effect is quite startling: I quickly began treating the keys like physical strings and playing it like a harp or guitar.
Instead of keys, what if you had small fretlike ridges or tongues or nubs that you pluck to create notes?
Sell kids a scaled-town toy model to learn on and then sell teens a full-sized one: the Motherplucker.
LINEAR KEYBOARDS: SUMMARY
I see no theoretical reason one couldn't play the standard piano repertoire on dual- or triple-offset chromatic keyboards: they conserve the left and right hand parts, though the widest intervals would be fingered a bit differently, using the spare board(s). But translating the repertoire would be a lot easier than from piano to guitar, and offset synths would be a lot easier to play than either current instrument.
Linear designs can be largely based on existing keyboard engineering, though they'd cost more than a piano-type synth, due to the duplication of keys. But a proper arrangement of keys is just as important as proper size, weighting, and feel. No serious musician will settle for a guitar or keyboard that's awkward to play if a better one's available. And any of the chromatic offset designs proposed here would be more learnable and playable than the traditional layout. An instrument that costs twenty percent more but will save you years of learning is an obvious bargain. (Music is the only field I know where a learning curve of thousands of hours is tolerated--and failure blamed on the learner!)
Any arrangement that staggers keys, whether diatonically like a piano or in a simple zigzag or grid, can compress or wrap a chromatic line of keys till one hand can play wide intervals and rich chords. But all these designs have inherent key-signature preferences that have to be overcome.
DEVICE 5: THE ZIGZAG
Admittedly, two scales is simpler than twelve, but you still need to understand musical theory. It's ideal for already-skilled musicians--but they've often invested years in the existing finger patterns. That was the barrier Paul Jankó ran up against in 1900. Anyone skilled enough to appreciate its advantages had already learned the conventional pattern the hard way... and had to unlearn it.
DEVICE 6: THE FULL JANKO
Because of the extra places you can hit a note, chords always fit the shape of your hand and your fingers are never too short to reach. While the whole-tone business feels odd at first, any chord or run you learn is the same in any key as long as you start on the right note, which may mean switching to the nearest up or down row.
You can also reach wide intervals with one hand on a Jankó--even small hands can reach an octave or more and the average is probably a tenth. A great advantage for classical musicians in particular!
The disadvantages: a pre-electronic key with three "access points" (one on each row), was feasable, but just barely. I haven't gotten to play a real Jankó yet, but it's reported the action was very stiff and that the far top and bottom rows felt quite different, being too near or far from the key's hinge-point. So your tones could get unexpectedly loud or soft if you used all the rows.
Also, beginners couldn't noodle around by just sticking to one row and get the sweetness built into a piano's white-key scale. It's intimidating at first--having to make your own scales. Musicians who can, usually know conventional piano already.
However, an electronic Jankó keyboard immediately solves the mechanical and feel-problems. I'd build it with only four rows--two for the main sequence, and one alternate in each direction. Costlier and more complex than a standard keyboard but it's a tested design: first-rank concert pianists used it and swore by it!
The noodling problem remains. As with a simple zigzag, it's easy to go from so-so to advanced on a Jankó, but the very first steps can be intimidating, especially for a musician weak on theory. Still, it's the one alternate design we know for a fact can be played at a professional level.
DEVICE 7: THE SPRAWLING GRID
Typewriter-like grids share problems with zigzags. Assume that notes to the right are a semitone higher, while moving vertically one key jumps an octave. Great--it's simple and consistent. But as you go up a scale, eventually you reach the edge of the world and sail off... The fingering patterns break down. It's like the guitar's bottoming-out problem, but here it happens at the end of each row. Chords must then be inverted. Some key signatures and chord patterns are centerish ones, some are edgy... and much harder.
One solution: a grid with a lot of overlap at the edges. Build the grid with chromatic rows of 25 keys (two octaves) and make each row up only ONE octave higher than the row below. The tones on the right of the lower row are duplicated on the left of the upper row--and so on up the grid. If you center your playing in the heart of the grid you can play most chords and runs without "breaking" at the edges. When you need to go high or low, you move up or down instead of way out to the side--they're accent notes, not often centered on.
This design, being stacked or sliced not continuous, and having a periodicity based on multiples of 12, is harder to re-tune to other scales, though it could be done. With a 16-tone scale, for example, you'd have an 16-key octave in the middle with only four notes above and below it, so "spilling over" would be more likely to occur. Scales with more than 24 notes become an obvious problem. Still, for the average user such concerns are nonexistent.
This may be a viable design, especially for non-classical players who don't make a lot of extended runs. A grid would be quite playable and the learning speed should be fantastic. The wastefulness of all those duplicate keys is offset by the compact arrangement: to get a piano's range you'd need something like 176 keys, which sounds cumbersome, but BECAUSE they're a grid, they all pack neatly into a briefcase-size array. Compact.
Maybe too compact! Two-handed playing could get cramped--one hand directly over the other, or crossing very closely an octave below. This may be why such designs haven't caught on. It's an obvious structure and must have been tried before. Why have I never seen one? Not even a historical footnote like the Jankó piano.
The two main radial designs are snails (conical spirals) and nautiloids (flat spirals). Concentric rings won't work, as there's a point on any ring where the rising notes have to drop back--a near-octave jump.
Spirals are the solution. To go up an octave you always move one key inward, but if you keep going around and around, you always rise a semitone per key. Such designs are fully chromatic, continuous, and compact, conserving the shape of a fingering pattern all over the keyboard.
But there's a heavy price: the keys in the center must be smaller to fit them all in. So both finger-spacing and the orientation of one's hand must change as one plays higher or lower. Hell on your wrists as well as your brain! It's too bad: right back to Pythagoras, logarithmic spirals and musical scales have been flirting with each other. There's a Turkish version of the oud that's quite radial/spiral, and of course the spirals of some larger horns, and of the cochlea inside the human ear... but aside from these, the love affair between spirals and music has stayed mostly theoretical. I'm not sure why. Are they just too disorienting for players? I think someone will eventually create a brilliant spiral design--spirals should be very ergonomic, since, after all, our hands swing in arcs--but all I can come up with is:
DEVICE 8: THE SNAIL
Picture a roulette table, with a big spinning cone instead of a crater. The cone has a spiral of keys (unequal-sized, the central ones smaller, so octaves line up and radiate from the center). The cone must be big, so the smaller treble keys near the apex are still playable). Ideally you could turn the cone with foot pedals instead of constantly crunching your wrists. Picture a light, quickly reversible pottery wheel. I bet music from such an instrument would have a fascinating character, but it'd be cumbersome--the size of a large desk. On the other hand, an acoustic piano rivals a Volkswagen bug...
Come on, someone think of a smaller verson.
A lampshade you wear on your head like your partyin' Uncle Ernie! You play your hat--how cool! (Ow, sore shoulders, fried hair, brain tumors! Don't sue me, don't sue me!)
Back to the drawing board. Maybe a mushroom shape like a Tiffany floorlamp, only spun with a pedal. A sort of large cone-shell on a stick, or a large collapsed umbrella. Which approaches my final design:
DEVICE 9: THE COIL
The helix is the only design that has a continuous scale, AND the compactness of wrapping, AND uniform key-size, AND a single fingering pattern for any scale or chord. Imagine winding a chromatic piano round a guitar neck (though neither piano nor guitar were my inspiration--it was the valiha, a cylindrical harp in Madagascar). I've experimented with coils wrapping around once per octave (12 keys), per fifth (7), and per fourth (5). I find the 12-key spiral no worse ergonomically than many existing instruments, though the cylinder must be rather fat to make the keys large enough to play, and most chords have the third, whether major or minor, in a rather awkward position. But guitarlike inversions, with the third either a couple of octaves higher or lower, aren't bad at all. Certainly it's no worse than a guitar fretboard, and vastly simpler to learn.
Over all, I prefer coils that repeat every perfect 5th, because the fingering for most common chords and scales lines up better than for octaves. So, in my preferred design, one turn of the coil has seven keys each a semitone apart, and the eighth key, directly above the first, is a perfect fifth above it. Unlike a guitar's strings, where strings are separated mostly but not entirely by fourths, this pattern of fifths repeats consistently all the way up the scale.
A vertical run takes you effortlessly through the circle of fifths.
A diagonal climb to the left, hitting every other key, is a succession of octaves.
Grasping the cylinder naturally creates major and minor chords with two fingers in the back playing a fifth and the thumb in front playing the major or minor third.
Comma-shaped patterns of adjacent keys form compact, easy inversions of both major and minor chords; major and minor sevenths have notes very close and lined up conveniently.
Major scales are particularly easy: the first four notes, say C D E F, step around one coil; the last four notes, G A B C, are directly above the first four notes you played! You just slide your fingers up.
The various minor scales are nearly as simple; you adjust the second "turn" to make it harmonic melodic natural or whatever. A beginner who blindly goes on sliding fingers up one note because it worked for major scales, will get a "minor mixolydian" scale, incorrect for most classical pieces but quite nice for folk or blues, and an easy base from which to learn the other minors.
In short, the coil is as idiot-proof as a serious instrument can be. In one day, you can learn to play it well enough to jam. And the structuring it imposes, unlike pianos or guitars, doesn't lock you into any key preferences; you could jam with horn players, guitarists, keyboardists, and singers, playing unusual chord jumps, chromatic passages, key changes. Whatever chords they name, you could play.
After one day.
My current mock-up has the 'vertical" lines of fifths slightly back-slanted, which moves all thirds and sixths slightly closer together. Like the stacked fifths, thirds and sixths are now also playable with one (diagonal) finger, rather like a bar chord on a guitar, but much simpler. The backslant also moves the more discordant tritone intervals slightly further apart and harder to hit accidentally, though they're still perfectly playable for seventh chords.
The cylinder can either stand on a table or desk with a rubber traction-foot. Or it can angle out from the body, with a guitar strap at the top and the foot attaching to a waist belt. (This is quite comfortable but the coil then looks as comically phallic as a straddled Congolese drum. This should please certain adolescent musicians). Leaning the instrument, putting pressure on the "foot" and bending its "ankle" could trigger tone-bending (suggested by sound engineer Lance Nottle of The Music Annex). Both the foot and the strap-mount on top can rotate, so the cylinder can be spun round or adjusted by the heel of either hand while playing. This eases quick chord shifts that don't fall into a simple I-IV-V chord pattern.
Longitudinal flanges or ridges about 1/8" tall stand between each pair of keys a semitone apart. These let the player cradle the instrument, turn it, etc., without hitting keys accidentally--any instrument needs SOME non-key areas to grasp! The ridges also prevent the accidental production of tone clusters due to sloppy fingering--one fingertip can't hit two (horizontally) adjacent keys as you can on a piano. The flanges are also tactile guides to the back side of the coil, where you can't see.
Due to the coiling effect, my prototypes (with 96+ keys, more than a concert grand) are still extremely compact--about two feet tall and two inches thick (60 cm x 6cm).
For now the coil will probably need an external amp/speaker, as it's unlikely that such a compact design could accomodate today's bulky batteries and speakers (though an earphone jack or a mini-speaker for practicing is feasible now, and a self-contained unit is just a matter of time, as strong rechargeable batteries and decent speakers get smaller. Even a prototype with an external amp will be more portable than an electric guitar.
Since pitches can't be distinguished by an irregular pattern of keys (as on a piano), the keys on my mockups are color-coded (labeled too--I believe in redundancy.) The resulting rainbow pattern is rather like the scales of a large tropical snake; it looks both high-tech and tribal, which can't hurt its marketability. The keys are all the same size and shape of course--lozenges about an inch wide and an inch and a half tall. Quite a lot like large snake-scales, in fact.
Perhaps "home" keys should have tactile markers as well. If each C is marked with a small bump and each A with a pit, they'll form a spiral grid such that all other notes are adjacent to a C or an A in SOME direction.
The external amp/speaker should probably contain the transformer and effects pedals too. Possibly even the synthesizer's control window for choosing voices could be here, if the coil itself is too cramped to allow it. In fact, any off-the-shelf synthesizer could do these amp/effects/voice-choice jobs, as long as the coil generates MIDI as well as a straight audio signal.
If any sound-patch controls can be fit into the coil itself, I'd do it as follows: have one button that turns the keys into sound-patch buttons for one keystroke. That is, push this button and the next key you hit controls the tone-bank--you'd have a hundred voices one keystroke away, and they're marked and color-coded already. You can even label each key with its resident instrument. Saves on buttons and is faster than entering tone-bank numbers.
I'd also have four to six buttons that store a player's favorite voices, so you can switch in performance on the fly. I find keyboards that don't have this (that is, most of them) annoying. I rarely use most sounds--why not have the useful handful be instant and let the rest be an extra button or two away?
To restate the obvious: this type of coil lets you play ANY perfect fifth with two adjacent fingers, or even one sprawling across two keys. It's always the same pattern--no adjustments for tricky key signatures like the piano's B or Bb. And exactly the same is true of ALL chords and scales. Every single major, minor and seventh chord can be learned the first day (in multiple inversions, too), since all major chords are fingered exactly the same, all minors the same, and so on. The I-IV-V pattern of chord changes (by far the commonest in music) can be played instantly by sliding the major-chord pattern one coil up for the V chord and/or one down for the IV chord--simpler than the C-F-G chords on a piano, and on the coil this pattern holds for ALL keys: A-D-E, E-A-B, you name it. It's as if every song in the world were instantly transcribed to C for the piano--and somehow, despite chord changes within a piece, miraculously STAYED in C. On a piano, to play major, minor, seventh and minor seventh chords in any key, you must learn forty-eight finger patterns (some of which are the same, but you still must learn which ones to use where). On the coil, you learn four.
Inversions of chords (different finger-patterns creating versions of a chord in which different notes predominate, cluster at the base, stand out at the top, or spread out across two or more octaves) are richer and easier on the coil than the piano or guitar. The range or "fatness" of even a one-handed chord on the coil is more guitar-like than piano-like, but inversions can be created by adding notes to the bottom as well as the top, which is more piano-like: guitars achieve many inversions by subtraction, i.e., muting strings. And some guitar chords have very few comfortable inversions. On the coil you always have many.
The tight coiling allows one hand to play quite large intervals--even a child's hand can span over an octave and an adult, two, three, or even four octaves. Not every single interval this wide would be playable one-handed, since some notes on are opposite sides of the spiral, but many are--all the octaves and most high/low fifths and thirds, for example.
Not all keys are visible at once. Even for beginners hunting for notes, this isn't insurmountable: some KEYS are hidden, but every note is visible and easily playable in SOME octave above or below the note hidden on the back side of the coil. Several easy fingerings exist for major, minor and seventh chords that have all the keys nearly in line vertically, so they can all be seen at once. Still, the temptation will exist to learn to play by sight, using only half the keyboard at a time, when many of the best fingerings involve grasping the cylinder and pressing keys back AND front. In the long run, the helix is best played by feel, not by eye. But then, that's true of most instruments.
Spirals are fascinating, but intimidating. People who've tried prototype coils find them strange-looking. We think in straight lines and grids. These patterns have dominated since the Industrial Age began, and the coil's advantages are distinctly nonlinear. It takes a few minutes of noodling to see how easily the keys fall in place under your fingers.
Like a guitar, playing a coil requires both mental agility and coordination of hands. Some notes are easier to reach around to with the left hand, some with the right--though WHICH notes can be arbitrarily adjusted by spinning the cylinder. Scales and long runs are playable with one hand but easier with two alternating hands when a key is on the left or right side of the coil. Of course, one-handed scales are hard on the piano too, and playing a guitar always takes two hands. But the coil really does use both hands equally. I'm fairly ambidextrous so I didn't think much about it, but players with one hand strongly dominant might find it hard to play fast runs, though chords should be easy for all.
THE ENGINEERING CHALLENGES
I'm not yet sure of the ideal size for the cylinder and the keys. The smallest model I've made, a bit over 2" wide, is the best so far. I'd like to go smaller still but I worry about cramming the circuitry into a tube much tighter than this. On the other hand the coil doesn't really have to have much of a synth's circuitry; it's only the input device. If it's ergonomically necessary for the tube to be smaller than the circuitry can fit in, the tube can be extended upwards or downward and the keyless space used to house the extra electronics. Effects knobs and possibly the voice selector could be in this keyless area.
Connecting with the power supply and sound-generator may be difficult. The cylinder will be far easier to play if it's freely turnable, but any electrical cords coming off an instrument freely spun will eventually get tangled. Even if they emerge from the axis, as a big coax cable, twisting is a problem. I can easily picture a bulls-eye-striped socket that could provide a modest number of always-separate in/out lines without twisting or tangling, but since it must be right at the axis, space will be limited. Cordless would be far better.
Decisions must be made about the optimal size and proportions of the keys, and how they should operate. Pressure sensitivity is a must. I'd prefer the ability to bend individual notes as well, by pushing a key hard toward the head or foot of the cylinder. Keyboardists live with a bend knob affecting all notes, and pianists make do completely without. But why should a coilist? Guitarists don't, and bending adds tremendously to the expressivity of an instrument.
It won't be possible to determine the optimal height of the ridges separating semitone keys until a playable model is built. They need to be high enough to allow easy cradling of the instrument without triggering keys, yet not so tall they impede playing or gouge the hand. I'd guess 1-2 mm high, a similar width, and 2-3 cm long. Clearly they should be rounded.
At present, I consider the most viable candidates to be the coil and the twin or triple offset keyboard (probably offset a fifth). Slightly lower priority would be a plucked variant of the offset keyboard, and a simplified Jankó design (with perhaps four rows), and the 25-key version of the sprawling grid. The radial and straight/zigzag chromatic keyboards are presently low priority.
But this may shift if I unearth relevant historical designs and find they were viable--or proven not.
FUNDING AND MARKETING
I haven't addressed funding or marketing at all here. Having a good instrument, preferably several good instruments, seems more important at this stage. I'm different from the other music inventors I know of, in that all of them are obsessed with cool sounds--with output. I think input is vastly neglected, and the most bang for your buck lies hidden there. Today, you can always get better sound patches or build an output device, smaller, faster, cheaper (and, unfortunately, louder).
The real challenge is to design controls that people 1) CAN play and 2) WANT to play.
Imagine the triumph of high tech in another problematic arena. Let's say cars are now clean, cheap, quiet, electric, and solar-powered--but you steer them with the same rudder used on a Model A Ford and stop them by throwing an anchor overboard.
That's an electric guitar. That's a synth. That's where we are.
It's a stupid place to be.
That's my marketing plan.
Chris Wayan, 2001
Calling all engineers, synth-building companies, inventors, tinkerers, even cranks:
I need you! I can design these, but I lack the resources to print the boards, mold the bodies, link the switches... If you like any of these designs enough to build it, let me know! My focus is on getting these actually built and playing them, not on potential profit; if you or your company can build and market any of these, you'll find me eager to help and not terribly possessive of the designs.
World Dream Bank homepage - Art gallery - New stuff - Introductory sampler, best dreams, best art - On dreamwork - Books
Indexes: Subject - Author - Date - Names - Places - Art media/styles
Titles: A - B - C - D - E - F - G - H - IJ - KL - M - NO - PQ - R - Sa-Sh - Si-Sz - T - UV - WXYZ
Email: firstname.lastname@example.org - Catalog of art, books, CDs - Behind the Curtain: FAQs, bio, site map - Kindred sites