What’s in this article: An overview of different ways of creating friction vibrations, a discussion of principles and techniques pertaining to the making of friction instruments, and short descriptions of several friction instruments I’ve made.
In recent years I’ve spent some time designing and building musical instruments that work by friction. The results have been mixed. Many seemingly promising friction ideas have turned out in practice not to be slam dunks. But a few have been more successful.
One should always define one’s terms: By friction instruments I mean instruments in which the initial vibration is excited by friction in one form or another, and by the word friction I refer to anything which can be seen as a “stick-slip” action. Stick-slips come about when two surfaces are sliding across one another. In such situations, it’s often the case that the slippage does not occur steadily and smoothly, but rather in a rapid series of tiny jumps: the surfaces momentarily grab and then, when pressure has built up, release and jump forward a tiny amount, then catch again and repeat. If the stick-slip frequency is in the audible range, then you’ll hear the sound. Many everyday sounds come about this way, from door-hinge squeaks to violin tones. A bit of oil on the door hinge lubricates the surfaces and prevents those momentary catches. A bit of oil on the violin bow would render it ineffectual for its intended purpose; better in that case to apply something to increase grabbiness, like bow rosin.
I got into friction because I was seeking certain kinds of sounds and some kind of instrument-tactility that I hadn’t quite achieved in other instruments I had made, and a friction approach seemed to hold some promise. I sought a particular sound quality that I was hearing in my ear’s imagination; a slightly voice-like, slightly thick, slightly complex, melancholy, wobbly, expressive, humanly imperfect, large-mammal kind of sound. Somehow it seemed like it might be possible to coax something like this from the old stick-slip. That said, I was also of course open to other sorts of unique and interesting sounds I might stumble into through friction explorations.
There are lots of ways to induce friction, and I tried as many as I could think of. Here are a few of the basic possibilities:
Bowing. This is a matter of drawing something across the main vibrating body. That main vibrator may be a string or something else. For the thing that is to be drawn across the main vibrator, violin bows, cello bows and the like use strands of horsehair. The hairs are held slightly taut by a bowed stick; in some non-western instruments the degree of tautness is controlled by fingers of the bowing hand pressing on the hairs. Many variations are possible; for instance, inexpensive bows may use strands of other materials such as nylon. Some instruments replace the bow with a rotating wheel with suitable edge for string friction. Some others have used a looped strand of string, fabric, or other flexible rope-like material, mounted on something like pulleys and continuously rotating past and across a string or other vibrating body. It’s also possible, in some cases, to hold the ends of a strand of string or fabric, and draw it back and forth across the vibrating body in a motion a little like someone drying their back with a towel. But, having listed these various possibilities, I should say that in most cases, for dependable, controllable and good-sounding friction sounds, it’s hard to beat a good old horse-hair violin bow with a bit of bow rosin. You needn’t assume that violin bows are to be used only on violin strings: horsehair bows are often the best choice for any number of other bowable objects, from cymbals to rigid metal rods to marimba bars.
Friction mallets. The usual friction mallet is a superball on the end of a stick. It’s amazing the sounds you can coax out of diverse sorts of flat surfaces using these things. Try it on windows and doors, refrigerator doors, various sorts of drumhead. The most promising candidates for the superball treatment are usually these sorts of thin, flat, diaphragm-like surfaces. For things to work well the mallet handle must be somewhat flexible, not rigid, and the superball needs to be attached well enough that it won’t just pull off. It’s usually best to hold the handle fairly near the superball, drawing the superball along the chosen surface. With practice you can learn to control pressure and speed so as to bring out the best sounds.
Direct finger friction. The idea here is simply to draw a finger or two across the main vibrating body to induce the vibration. The vibrating bodies that work well with this are usually things that are more massive and rigid than strings — for instance, this sort of technique has been used with metal bars or rods, various glass objects such as wine glasses, and cowhide conga drum heads. Some kind of friction enhancer is usually needed, and in many cases water works well – for instance, you need to wet the finger before stroking the wine glass. We’ll be talking more about friction enhancers later in this post. [Added note: here’s an example I just thought of where the main vibrating body is not heavier than strings, contrary to what I said above: balloons! As many children have discovered, finger friction on balloons is an excellent way to make interestingly irritating sounds.]
Using an intermediate friction-inducer. If you’re familiar with the Brazilian friction drum known as cuica, you’ll have some sense of this. In the cuica, the player strokes a small stick which is attached to a drum head. The stick-slip happens in the stick-stroking, and the vibration is transmitted through the stick to the drum head. (In this case, the player uses a moistened cloth, pinching the stick between the fingers with the cloth in between, as the best way to get a good friction response.) An overly simple analysis would say that the main vibrating body is the drum head, and the stick is there merely to deliver the friction. The reality is a little more complex: the stick and the drum head actually form a coupled system, with the stick contributing to the overall mass, which affects frequency. For another quasi-separate friction inducer, consider the Baschet Crystal, created by the Baschet brothers in France in the 1960s. Here a set of glass rods, stroked with wetted fingers, deliver the vibration to heavier tuned steel rods. Alternatively, rather than using rigid materials, you can use strings as friction-inducers. For instance, strings have been used with drum heads in a manner similar to the stick in a cuica. I’ve explored stroked strings attached to other vibrating bodies; for instance very light-weight friction strings attached to much heavier musical strings, as well as friction strings attached to rigid materials like kalimba tines and free bars. An interesting variation on the stroked string idea which you may have seen: tie a loose loop at one end of a string around a rosined stick; affix the other end to the head of a very small drum, and whirl. The stick-slip friction that arises where the loop rides over the rosined stick will be transmitted through the string to the drum head, and you’ll hear the tone coming from the drum. (Be sure to make a shallow groove in the stick to keep the loop from flying off the end.)
Drawing the main vibrating body across a friction-inducing surface. I can think of just a few cases where something like this happens; perhaps others can think of more. Here’s an example: Start with a 2×4 board of between, say, 4 and 8 feet long. Standing on a wooden floor, hold the board in front of you so that the lower end is on the floor, with the board leaning back slightly toward you. Keeping that angle, push the board along the floor, so that the lower end skitters across the wood. If conditions are just right, the board will start sounding at one of its resonant frequencies. There are other ways to set up this kind of system, and the San Francisco builder Tom Nunn has been one of the main people to explore them. His Lukey Tubes involve cardboard tubes, typically about 2-3 feet long, pushed end-down across the surface of a stainless steel sheet resting on balloons. Different lengths and thicknesses of tubes, and different speeds and pressures, all affect the pitch and tone quality. The sound is quite delicious. Hard to control though, Tom reports.
Wood and pewter, squeaky hinges. Here are a couple more friction-sounds that don’t quite fit any of the descriptions given above. There’s a kind of bird call that uses a small piece of pewter shaped so that it can fit into an opening drilled in a small, hand-held wooden piece. Twisting the pewter in its wooden cavity produces a nice bird-like sound. And then there are hinge squeaks, as in squeaky doors. I like these sounds and I’ve spent some time thinking about whether there are ways to harness them, manipulate them for interesting, variable, musical or quasi-musical sound qualities, perhaps even to manage the resulting pitch. No big breakthroughs so far.
Those, then, are some of the ways one can approach friction sounds. I’ll now mention a few things I’ve learned about friction sounds and how to make them work.
A key to managing friction sounds is finding ways to impose some discipline on the stick-slip and its frequency. Example: In the case of the violin bow, control of the sound can be easily maintained. The natural frequency of the string dominates the whole system, as the string imposes its frequency on the stick-slip events happening where the bow crosses the string. True, with beginning violinists you often get unexpected squeals, or higher string modes imposing themselves in ways that are not wanted, but for the most part, with even a minimally competent player, we get the intended tones in a predictable and controllable way. By way of contrast, think now of a door squeak. Here the situation is much more fluid: typically the stick-slip frequency varies quite a bit with the result that the pitch bends up or down,abruptly jumps from one pitch region to another, changes its sound quality, or stops and starts unpredictably. There’s no obvious way to impose discipline on the system.
So what’s the difference between the bowed violin and the door hinge that makes the violin more manageable? The key factor, I think, is that fact that the violin string has a strong and well defined resonance frequency that it successfully imposes on the stick-slip action. Essential to this ability to impose is the fact that the bow hairs are light in weight, and quite unresonant themselves, making them more likely to accommodate the demands of the more willful string. In the hinge squeak, it’s less likely that there would be a single strong and well defined natural frequency on either side of the hinge. (But note that the door itself makes a good soundboard. This may not help with frequency definition but in other respects it helps to make a nice sounding system.) The upshot is: if you’re thinking about creating a friction instrument with predictable pitch, you’ll need to be sure that one side of the stick-slip or the other has a strong and well-defined resonance peak at the desired frequency, and that that side is of sufficiently high impedance – that is, is massive and solid enough relative to the other side – to dominate the system. Or, to be more sure, you can ask yourself if there are ways to put both sides into frequency agreement.
Here’s another consideration that’s easy to overlook. In the case of the violin and bow just cited, the string under the bow produces the same frequency as the string alone. That’s not always the case: more often, the distinction between the friction-inducer and the main resonant body isn’t as clear-cut as it is with bow and string. Instead, the two elements work together as a coupled system, and their combined mass affects the resulting stick-slip frequency. As an example, consider again one of the really wonderful friction instruments, the Baschet Crystal. In this instrument, as mentioned above, a steel rod, rigidly mounted at one end, has a lighter glass rod attached perpendicular somewhere along the steel rod’s length. Sound is produced by stroking the glass rod with wetted fingers. The instrument consists of a set of these glass-and-steel assemblies plus a sheet-metal sound-radiating surface. In some descriptions, it’s suggested that the frequency is determined by the steel rod, while the glass rod functions solely as a friction inducer, analogous to the violin bow. But in fact the two rods behave as a coupled system, with their combined mass contributing to frequency. The position of the glass rod along the steel one also affects frequency, in part because the distribution of mass has leverage effects relative to the mounting at one end. This situation comes up frequently: the thing you might be inclined to think of as merely a friction-inducer actually works as part of the vibrating system and influences the whole. Horsehair bows, with their minimal influence on frequency, are the unusual case.
When thinking about friction instruments, it’s good to think about the direction of the impulse. Once again, our clear-case example: the violin string vibrates side-to-side, and the bow too moves side-to-side. The bow movement is in agreement with the direction of movement of the intended vibration, and all is well. Contrary example: When I said to myself: “I wonder if you could get a kalimba tine vibrating by friction,” my first thought was simply to draw a rosined finger along the length of the tine. This produced no result, because the direction of the impulse was not the direction in which the tine naturally vibrates. Eventually I made an instrument in which the ends of the tines are bent at a right angle. The player strokes this end-portion, so that the stroking direction now corresponds to the tine’s natural direction of vibration, and the tines speak readily. Wine glasses played by a rotational friction motion are an interestingly ambiguous case: the vibratory patterns of the glass, which correspond to modes that are well known from the study of bells, are mostly perpendicular to the finger motion: at any given point the rim moves toward and away from the center while the finger does its circumferential motion. But one could imagine that, due to the circular form of the rim, rotational stroking along one part of the rim would translate to a movement in a more “usable” direction of impulse elsewhere in the rim. One more example: While musical strings normally vibrate transversely – that is, side-to-side, as when the string is bowed – it’s also possible to set up longitudinal vibrations in strings. In this case, the vibratory movement is along the length of the string, with pressure wave fronts running back and forth from end to end. In keeping with the “direction of vibration” idea, you can induce such a vibration by stroking the string longways with rosined fingers. If you wish to try this, keep in mind that the speed of wave travel in the hardened steel of a typical musical string is extremely high. This means that the resulting frequency – which corresponds to the time it takes for a pressure wave front to make one round-trip excursion, end-to-end, in the material of the string – will be so high as to be just about inaudible for a string of typical length. To get a recognizable pitch you’ll need a string of 10 or 12 feet for a very very high note; 50 or 60 feet to get closer to midrange, and a hundred and more for low notes.
Most friction techniques require some kind of friction enhancer. For violin bows, it’s rosin (or resin), which is dried tree sap. You can buy rosin from a music dealer or borrow it from a local pine tree. It can be used in the form of a chunk of the stuff which can be rubbed on bow hairs or other friction surfaces, or it can be used as a powder (if starting with a chunk, you can powder by hammering). If you prefer a liquid form, the powder can also be mixed with alcohol to make a soup that you can spread on a surface or dip something in. You can also purchase rosin less expensiveley in quantity in the form of copal, normally used as incense. Copal comes in lumpy form, which can be hammered into powder. There are grip-enhancer sprays too, made for athletes, which seem to have rosin as a primary ingredient. I’ve found that rosin in one form or another seems to be the best bet for many friction situations, certainly not only violins.
But not all. As we can see from the wine glasses example, plain water works best in some cases. Some suggest a drop of vinegar in the water, and there’s a bit of conventional wisdom to the effect that hard water works better than very soft water. Water is generally best for glass friction surfaces (and many have found glass, by the way, to be one of the best friction surfaces for musical purposes, witness the choice of glass rods for the Baschet Cristal). Water is also good for use with wooden sticks, as in cuica. It can also be used, I have found, on plastic rods. As illustrated by the cuica, a good method in many situations is to use a bit of wetted cloth. One trick is to wear a wetted cotton glove. (White cotton gloves are available in bulk, since they’re sometimes used in situations such as jewelry work.)
As I mentioned at the start of this article, my adventures in friction have yielded mixed results. One of the main problems I encountered was inconsistency. Violins are dependable and predictable; many other friction instruments are not! In explorations with friction instruments I often found that I was able to get good results one minute, only to find the next minute that the instrument was responding with some pitch other than the intended, or sailing off into some unwanted mode of vibration. Even worse, there were otherwise promising instruments that sometimes would simply refuse to speak. A terrible performance situation: the instrument that sounded so great in rehearsal is perversely unresponsive at the crucial moment. A word of advice: avoid food that is even the least bit greasy before attempting friction music in the presence of an audience.
As I mentioned earlier, I got going in the friction game because there were some sounds I had in my imagination that thought might be possible through the old stick-slip. I did have some luck with this; in particular, one instrument (described below) came very close to creating one of my most cherished imagined sounds. Unfortunately, that instrument was among those that proved seriously ill behaved when it came to pitch-dependability, readiness to speak, and so forth. But, equally important, I also stumbled my way into other cool and sometimes outrageously raunchy friction sounds that I hadn’t foreseen. Following here’s a sampling of friction instruments I’ve made. The descriptions here are very brief. For more information to their respective pages in the Instrumentarium section of this web site.
By the way, I’m not finished … I hope to make a few more friction instruments as time goes by.
This is the one that came close to producing one of my most fondly imagined friction sounds. As you can see by the name, it’s related to the Baschet instrument mentioned above, in that a stroked glass rod is the key friction component. It consists of a set of tuned air resonator tubes, in each of which one end is covered with a plastic diaphragm. In the center of the diaphragm the glass rod is mounted, extending out beyond the end of the tube. The glass rod and the diaphragm form a coupled vibrating system, which is tuned by a combination of two factors: carefully thinning the diaphragm to adjust its rigidity (less rigid = lower pitch), and weighting at the center (more weight = lower pitch). The diaphragm-and-rod assembly is thus tuned to agree with the resonance frequency of the attached air column, and it was my hope that with this mutual reinforcement in place, the pitch situation would be reasonably stable. Alas, that wasn’t the case – the thing suffers from pretty serious instability of pitch. This is probably at least in part because the tuning of the diaphragm by thinning is iffy – not very precise or well focused, distorted by seemingly trivial irregularities in how it’s attached to the resonator tube, and very much subject to small variations in temperature. But the tone, reinforced as it is with the fullness of the air resonance, is wonderful.
This thing uses very light-weight strings for friction attached perpendicular to much heavier sounding strings. When you stroke the friction strings with rosined fingers, it sets up a stick-slip at the frequency of the attached sounding strings. The result is consistent and dependable: the sounding strings successfully dominate the stick-slip frequency. I made this as a bass instrument because it really helps to have the sounding strings much heavier than the friction strings. With lighter, higher-pitched sounding strings, I’m less confident that the pitch would be dependable.
This instrument uses plastic rods sounded by finger friction using either wetted cotton gloves or powdered rosin. By stroking the rods longways, you set up a longitudinal vibration in the rods. The longitudinal vibration pattern is closely analogous to that which occurs in tubular wind instruments, and it’s remarkable how much this instrument does sound like a wind instrument such as recorder. But here’s a notable difference between this thing and a tubular wind: as discussed above, the speed of sound wave propagation in air is much slower than it is in solid materials. This means that to get a note of the same pitch, a longitudinally vibrating rod of solid material must be quite a bit longer that the corresponding wind instrument air column. The harder the material, the greater the length requirement. I chose this type of plastic for this instrument because it’s relatively soft compared to other possible rod materials, and the rate of wave propagation, while still faster than air, is a lot slower than it would be in, for example, steel rod. This means that the rods didn’t have to be ridiculously long to get notes in the musical range.
This instrument is very similar, in its function, to the Pliffers described above, being based upon rods which are stroked lengthwise to produce longitudinal vibrations. The rods in this case are made of aluminum. While aluminum is still much softer than a material like steel, it’s still considerably harder and has a considerably faster wave propagation speed than plastic. This means that the rods for dansal 3 must be longer that those of the pliffers, and even so the range of these notes is very high. Also, the aluminum has far less internal damping than the plastic, so while the tone of the plastic disappears the moment the stroking stops, the aluminum rods have a most impressive sustain, continuing long after the end of the stroke. The best way to sound these, I’ve found, is by stroking with fingers or gloves sprayed with grip-enhancer spray. A generous dose of powdered rosin will also do the trick.
This is the finger-stroked kalimba mentioned earlier in which the tops of the tines have a 90 degree bend which allows stroking in the direction required to that set up the vibration in the tine. It’s electric (very quiet when not turned on). The preferred friction inducer is the grip-enhancer spray, but powdered rosin also works.
I was experimenting with glass rods on various sorts of large tongues, hoping for a nice, warm sound, when I hit upon this configuration with its pig-snortingly awful sound. The sound seemed interesting enough to justify building a proper instrument out of it. In this instrument the glass stroking rods are actually hollow glass tubes. The large tines, made of spring-steel, have machine screws attached through holes near the ends and projecting upward, and the glass tubes are loosely placed over these. The glass rods are stroked with wetted fingers. The tines try valiantly to produce a clear tone, while the glass tubes rattle mercilessly against their mountings.
These are threaded rods mounted in circular steel plates, played with a horsehair bow. In principle, it’s not much different from the traditional (but rare) instrument known as a nail violin. Easier to tune, though, because you can adjust the active lengths of the rods by means of the hex nuts that secure it to the steel plates. The tone is a high whistle. The fundamental tone of the rods isn’t normally sounded, but the rods speak with a clear tone at a couple of different overtones. The initial challenge in playing is to get it to sound whichever overtone you’re after at that moment, rather than one of the others. With a little practice this becomes possible and the sounds become quite manageable.
In this one, strings for friction-stroking run from a yoke above to a set of small but very rigid spring-steel tines. The tine pitches are very high. When you stroke the strings with rosined fingers, you hear the tone of the tines, and you also hear a kind of moaning sound that is other stick-slip frequencies arising from the string-stroking. It’s an odd combination.
The main sounding elements here are rigid bands of spring-tempered steel, ranging from about 18″ to 36″ long, rigidly mounted at the lower end and extending upward. There are three ways to set them into vibration. One is the obvious non-friction method: you can bong them with a superball beater for nice gong tones. For friction sounds, you can stroke them with superball mallets for luscious moaning sounds. Or you can use the long monofilament nylon strings attached to each band near the base. These strings are just freely hanging there, looking innocent but having been rubbed with rosin in preparation for playing. Grasp one between thumb and forefinger and pull away from the instrument, allowing the string to slip through in a friction-inducing way. This produces a rich gong-like tone in the band, but with a sustaining rather than percussive. These friction sounds are fairly quiet in themselves, but pickups positioned near the bases of the bands allow amplification.
(I’ve also made various bowed strings, but they’re not listed here because they don’t break much new ground, friction-wise.)