This article is about mbiras, kalimbas, sansas, and similar musical instruments with vibrating tines or tongues, including non-traditional or home-made variations. These vibrating-tine instruments, in all their diversity, make up the family known as lamellaphones. Within that field, the article’s focus is on overtone-tuning - the practice of adjusting the overtones within an instrument’s sound to bring them into better alignment with the fundamental tone.

Overtone-tuning is often done with other sorts of instruments. For instance, people who are serious about marimba-making know that overtone-tuning, or its absence, is central to the character of that instrument. The same applies for some other tuned idiophones such as steel drums and carillon bells. But for some reason, the question of overtone tuning isn’t often considered when it comes to making lamellaphones. That’s too bad, because lamellaphones can benefit from overtone-tuning just those other instruments can, and it’s at least as easy to do.

This first half of the article lays out the foundational information for the process of overtone-tuning in lamellaphones. The second half of the article, to be posted here on the Experimental Musical Instruments news page someteime soon after this one, will provide the step-by-step how-to for two possible overtone tunings.

Overtone Tuning - the basic idea

[You may choose to skip this section if you already have a good conceptual grasp of harmonic and inharmonic overtones.]

While we think of a typical musical tone as a single note, most musical tones actually have several frequencies present in the sound. Usually the lowest of these frequencies, commonly called the fundamental, defines the pitch. In other words, the pitch of the overall sound, as perceived by the listener, is the pitch of the fundamental. Other frequencies present in the sound can be thought of as overtones. They contribute a lot to sound quality or tone color, while playing a lesser role in the perceived pitch.

The musical relationships between the fundamental and its overtones vary from one type of instrument to the next. For many musical instruments, the relationship is harmonic. In this context, “harmonic” does not simply mean “harmonious;” it refers to a set of specific mathematical relationships between the frequencies. In sounds containing harmonic overtones, the overtones seem to blend imperceptibly into the fundamental. The pitch-sense is very clear, and the tone seems coherent and integrated. Instruments with harmonic overtones include string instruments (assuming the strings are well made and under sufficient tension), and wind instruments (assuming the bore shape is suitably designed for the purpose).

Then there are instruments that produce inharmonic overtones. For these, the simple relationships of the harmonic series don’t apply, and overtones appear at various quirky pitches relative to the fundamental. This is the case for most drums and idiophones. With these inharmonic tone qualities, the pitch-sense tends to be less clear and the tone usually seems, subjectively speaking, less integrated. This isn’t necessarily a bad thing: inharmonic tone qualities can be piquant and colorful. The world would be a boring place if all musical sounds were somehow required to be perfectly harmonic.

Nonetheless, if you can shift those unruly inharmonic overtones into harmonic agreement with the fundamental, the clearer tone quality that results is often worth the effort. This is especially true when the instrument plays in dense and complex musical contexts. Prominent inharmonic overtones in such contexts can be musically confusing. In some cases, the listener’s ear may even tend to track the overtones rather than focusing on the intended fundamental pitches (definitely a confusing musical situation). With the instruments mentioned above - marimbas and xylophones, steel pan and similar instruments, bells, and the instruments of the lamellaphone family - it’s possible to realign at least some of the overtones, moving them from potentially confusing inharmonic relationships to much clearer harmonic relationships.

 Overtone Tuning in Lamellaphones - getting more specific

In lamellaphones, as in marimbas and other instruments, overtone tuning is especially valuable for notes in lower registers. In lower notes the overtones tend to come through strongly because they are near the heart of the audible musical range. Sometimes they even dominate the fundamental. Overtone tuning is less important in for high notes because the overtones in high notes, being much higher still, are less conspicuous. For this reason, it’s usually not necessary to tune the overtones for high notes. This is convenient, because the task of overtone tuning is more difficult and time-consuming for higher notes. How high before overtone tuning is unnecessary? This can be decided on a case-by-case basis.

With marimba bars, overtone tuning is done by reshaping the bars, and the same approach works for the vibrating tines of lamellaphones. The underlying idea is as follows:

The fundamental and the overtones in the sound correspond to different patterns of vibratory movement, or modes of vibration, in the tine. These different patterns have different regions of maximum flex over the length of the tine. If you thin the tine in the regions of maximum flex for any given mode of vibration, the resulting weakening will lower the frequency of that mode. It will have much less effect on other modes not sharing the same flex points. So if you can figure out where those regions are and do your selective thinning accordingly, you can adjust the relationships between the overtones and the fundamental, bringing them more in line with the ideal of harmonicity.

This process of selective thinning is potentially a complex and difficult task. It certainly is in marimba bars, where the tuning of multiple overtones can reach the level of high art. But in lamellaphone tines, it’s possible to achieve very nice results with a simple approach. The approach I’m going to suggest involves tuning just one overtone relative to the fundamental. The process is systematic, predictable and not labor intensive.

Why is it that you can get away with tuning just the fundamental plus one overtone? There are several reasons, but the main one is that in vibrating tongues the overtones are typically very widely spaced. In a uniformly shaped, rigidly mounted tine, for instance, the first overtone above the fundamental typically appears around two octaves and a minor sixth above the fundamental. That first overtone, despite being so high, is often conspicuous in the tone. For this reason, it makes sense to tune it, especially for notes in the lower registers. Additional overtones are arrayed higher still. These additional very high overtones are usually less conspicuous and tend not to affect the pitch-sense as much as the lower overtone. The ear tends to hear them just as added color. For that reason, tuning those very high overtones is less important.

With these things in mind, we’d like to tune the tine so that that first overtone over the fundamental - the one that stands out most significantly in the tone - ends up in some coherent, harmonic musical relationship to the fundamental. In the second installment of this article I’ll suggest two tunings. One positions the first overtone three octaves above the fundamental, and the other positions it two octaves and a fifth above. Both of these fall within the harmonic series and provide the kind of musical coherence we seek. My preference is for the octave tuning because it’s clearer and more lucid - I love the sound of it - but the quint tuning (the one using the fifth) offers a bit more personality.

So stay tuned. I’ll be posting the second installment here on the Experimental Musical Instruments news page within a few days of this one.

~~~~~end of part one~~~~~~~~~

Lately I’ve been experimenting with string instrument soundboards that are not attached the sides of the sound box all around the periphery, but are attached only at the ends.   The illustration below will give you the picture, using a rectangular sound as an easily visualized example.  I know I’m not the only person to explore this sort of design, but I can’t recall where I’ve seen it before. 

The idea is to allow the soundboard freer vibration than would be the case if it were rigidly attached all around.  This is in keeping with something I’ve long been interested in: string instrument soundboards that are unusually light and/or flexible, so that the strings can easily activate them and drive them to large amplitudes. Because the strings deliver their energy to such soundboards rapidly, the resulting sound tends to have relatively short sustain and the tone may be kind of punchy. For lower notes there may be a bit of a thump in the attack.  The tone tends to be strong in the fundamental, de-emphasizing higher overtones. In theory such soundboards should be louder than more rigid, longer-sustaining soundboards. I haven’t always found that to be the case, but greater volume isn’t the main reason I’m interested.  The main reason I’m interested is, I like the punchy, saturated tone.

A very weak soundboard is likely to distort or collapse under the pressure of the strings.  So what I’ve been doing with these semi-detached soundboards is placing supports of dense sponge rubber under the sides of the board at selected points. (This is the same dense sponge that we sell here at EMI.  It’s great stuff.)  This prevents the soundboard from collapsing under the pressure of the strings, yet it still allows high degree of flexibility. It also adds damping to the board. The damping is probably a good thing, evening out the response and reducing the likelihood of wolf tones (wolf tones = disproportionately loud notes resulting from pronounced resonances at certain frequencies). 

Soundbox with floating soundboard

So far I’ve built three box zithers this way and a guitar-like instrument, as you can see in the photos. The soundboards are redwood, about a tenth of an inch thick.  They have no strutting or ribs underneath, except for a little cross-grain strip for added strength under the bridge in the guitar. On all but one of the zithers I added magnetic pickups, custom made to accommodate the required pickup width. (I tested piezo pickups as well as little on-board microphones with these instruments, but the magnetic pickups sounded best to me.)

So how do they sound?  You can hear them in the samples below.  I totally love the sound of the zithers, both unamplified and amplified. They really do, to my ear, have the saturated, punchy sort of tone I was after.

The sounding results with the guitar suggest a rather different story. Unamplified, the guitar sounds pretty blah. The mechanics by which the strings drive the soundboard are quite different in zithers and guitars, and what worked so nicely for the zithers is noticeably less pleasing in the guitar. But when it’s played through the pickup, I like the tone more than that of a typical electrified guitar. To my ear it sounds a lot less electric and more acoustic even than an acoustic guitar with a soundhole pickup.  The main reason is the way the highly compliant soundboard interacts with the strings and colors the string sound that the pickup hears. In addition, a couple of other design features contribute to a more “acoustic” tone, namely: 1) I’ve strung the guitar with an unusual set of soft steel strings (more about this in another posting sometime in the future).  2) The pickup is positioned as far from the bridge as possible - adjacent to the 16^th fret, which is much farther from the bridge than is typical. This contributes to a warmer tone.  (If you look back at the zithers, you’ll notice that the pickups are right at mid-string, for a similarly warm sound.) 

(About the use of the word “floating:”  duclimer makers sometimes use the phrase “floating soundboard” to refer to a soundboard which isn’t affixed to the sides of the box, but held there only by string pressure.  I’m taking the floating idea a step further, with soundboards that don’t even touch the sides all around.  The term “floating bridge” is occasionally used for guitar bridges that stand on two feet rather than lying flat against the soundboard; one again I’m taking things a step further with a bridge that doesn’t touch the soundboard at all.)

A Pair of Floating Zithers. These two box zithers are made as a pair to be played together by one or two players. The ranges complement each other, providing three and a half octaves between them. With seven strings per octave, the tuning is diatonic, but sharping levers on the left allow for quick and convenient retuning to different scales. The weighted steel arm on the larger zither adds a vibrato to the tone. Give the weight a little push and it will bob up and down because of the way it's anchored in the strings. It continues to bob as you play, wavering the tension on the strings to create the vibrato. The weighted arm doesn't work as will on the smaller zither, so it has an unweighted arm that the player can flex to provide the vibrato manually.

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

Floating Guitar. This guitar has a rather poor tone unamplified, but the amplified tone is, to my ear, natural sounding and appealing. For more details see the description in the main text above.

Over-Under Zither. This zither is almost identical to the smaller of the zither pair above, but it has a mid-string bridge in the form of a weighted bar. The string segments on one side of the center bridge are twice as long as on the other side, so the notes on either side are an octave apart. The weighted center bridge isn't anchored to anything but the strings themselves. It's free to wobble and bob, creating pitch-shifting and tremolo effects in the strings.Because of the unfixed bridge, the tone is a little thin and plinky, and the tuning is pretty seriously unstable, but the peculiar wobbly effects appeal to my ear nonetheless.

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

 I’ve been making an effort lately to get useful information on piezo films into these pages for the benefit of anyone who may want to use them.  Here’s one more posting along those lines. The question I get most frequently from customers regarding the piezo film is, is it OK to trim a piezo film pickup to a smaller size?

The quick and safe answer to this question is “Beware: cutting a piezo film pickup may ruin it. But a more complete answer is, sometimes it’s OK to cut the film, sometimes it’s not; it depends on how you cut it. Read on for details.

The piezo film pickups we sell are made up of a very thin film of piezoelectric material protected by a layer of clear plastic lamination on both sides. It’s hard to see this unless you look quite carefully, but the piezo under the plastic is actually a double layer, in the form of a single piece of very thin piezo film folded back on itself. The two terminals for the wire connections, located alongside one another at one end of the pickup, are actually at opposite ends of this underlying piezo film, brought next to each other by the folding. (Think of a towel folded in half: the folding brings the opposite ends together.) Now imagine that you decide to shorten the pickup by snipping off the end. By cutting off the point where it folds, you are (unknowingly) cutting the piezo film into two separate sections; it’s then no longer a single sheet. The electrical continuity between the two terminals is broken, and the pickup will not function

So that’s the big negative: you can’t shorten the pickup by snipping off the end; that will destroy the pickup

However, you can trim in other ways, and as long as you do so without compromising the electrical continuity, the pickup will still function. First of all, you can always trim off the clear plastic borders; there’ll be no ill effects because there’s no piezo material in those clear portions anyway. Second of all, if you want a narrower pickup, you can usually get away with trimming the sides, as long as you make sure that the folded end isn’t snipped off

Keep in mind, too, that the piezo film pickups are quite sturdy. If it’s necessary to give the pickup a sharp bend or something of a fold to make it fit in the intended application, that’s ok.

A couple of months ago I did a posting titled “Wiring Multiple Piezos Together.”  That post has information on how many piezos you can run to a single output before you start to lose signal strength and fidelity, followed by suggestions on how to wire such hook-ups.  The other day it occured to me that I didn’t address one important aspect of the topic.     

What I neglected to discuss was phase cancellation.  In multi-piezo installations where the piezos are on  independent sounding bodies — say, different bars on a marimba – then phase relationships aren’t an issue. But when you have two or more piezos on the same vibrating body — as, for instance, on a soundboard – it’s likely that the signals coming from the piezos will not be perfectly in phase. Then there will be some cancellation when you join them, resulting in a loss of signal strength. Depending on circumstances, the cancellation could be negligible or serious.  

The rest of this post discusses this issue.  To simplify the discussion, I’ll use the case of two piezos mounted on a soundboard as an example.   

There are two ways cancellation can come about when two piezos are mounted on a soundboard: 

1) Phase reversal.  Something about the way the piezos are positioned or wired could result in the piezos capturing opposite phases of the soundboard’s vibration.   This could happen for any of the following reasons: a) one piezo is positioned on the inside of the soundboard; the other on the outside.   b) One piezo is reverse-wired relative to the other. c) One piezo is attached upside-down relative to the other – that is, the two piezos are  mounted with opposite sides against the soundboard.  In any of  these cases, if the piezos are located reasonably close to each other, then their two signals will be close to 180 degrees out of phase, and serious cancellation will occur.

2) Phase offsetting. If the two piezos are somewhat far apart on the soundboard, then each will, at any given instant, be picking up different phases of the waves as they travel through the soundboard.  The phase difference will most likely be much less than 180 degrees, so cancellation will not be severe.  More interestingly, different partials within the sound (with their different wavelengths in the soundboard) will have different and variable phase relationships at the pickup locations.  Keep in mind that this kind of effect occurs naturally in the acoustic sound coming off the soundboard anyway; it’s part of the sound our ears are used to hearing and not necessarily a bad thing.  In spite of being less efficient, with luck it will contribute to a sound that is warmer and more natural than what a single pickup would capture. 

Practically speaking, what can you do to manage these phasing questions in multiple piezo hookups?

First, regarding the 180 degree cancellation described in #1 above: This is normally an undesireable situation and it’s easily remedied by reconfiguring things so that the two piezos are in phase.  One easy way to correct an inverted phase situation is simply to turn one of the piezos over so that its other side presses on the soundboard.  On the other hand, if you’re interested in a weird hollowed-out sort of sound and are willing to accept a serious loss in signal strength, you might enjoy exploring the possibilities of a deliberately out-of-phase hook up.

Regarding the subtler phase interactions described in #2 above:  In theory you could try to analyize oscillation patterns and their interactions as a way of determinining ideal soundboard locations for two or more piezo pickups. In practice this isn’t very realistic, given the almost infinite complexity of possible tones and all their harmonic and inharmonic partials in interaction. Instead, it’s usually worthwhile to spend a little time in trial-and-error mode, testing different locations for the piezos, including varied distances apart, in search of the most attractive sound. 

In most cases and for most musical tastes you’ll find that having two or more piezos on the same soundboard can yield a warmer and more satisfying sound than a single piezo would.  An exception is the case where you want a particularly bright and edgy tone with a sharp attack. That’s because the cancellation with multiple piezos happens most in the higher frequencies/shorter wavelenghts (as is the case in the natural acoustic sound as well).  A single piezo, being free of these high frequency cancellation effects, tends to sound brighter.  Also, multiple piezos do a better job of capturing the complexity, variablility and non-static nature of the natural sound. This makes for a softer attack,  and contributes to what the our musical ears hear as a warmer and more alive sort of sound. A single piezo, by contrast, will tend to have a slightly more “clinical” feel to its sound.   

We sell piezo material in several forms for making musical instrument pickups. One form is cable, about a tenth of an inch in diameter. The piezo cable looks very much like regular shielded audio cable, but it’s got a very thin layer of piezo material inside which enables it to function as a contact pickup. It’s most often used in or under the bridge of string instruments, where it can respond very directly to the string vibrations.

For a typical guitar installation, you need a piece of cable about 5″ long to span the bridge with enough left over to make the necessary connections. But our customers have occasionally wondered: how much longer can a piece of piezo cable be, and still work well?  Imagine for example that you’re making something like a clavichord or harpsichord; can you use a single piece of piezo cable several feet long to pick up the sound from the full length of the bridge?

I recently performed some tests to see how length affects the performance of the cable. Mind you, the tests were very rudimentary. I didn’t build harpsichords to test them in; I just cut pieces of the cable to various lengths and hooked them up to an amplifier. I tested and compared their outputs by wedging them under the strings of a guitar and plucking the strings.  Here’s what I found. 

Finding number one: Length doesn’t seem to be a problem. Long cables were about as responsive as short ones.  The longest I tested was 48″, and the strength and quality of the response was not very different from that of a 5″ cable.

Finding number two: Regardless of length, the cables do not produce a very strong signal. Theoutput is less than that from the piezo films that we sell.  (This wasn’t really news, but these tests confirmed it for me.) That leads to these two pieces of advice: 1) Piezo cable is best used in applications where it can be given strong vibrations to respond to (such as in-bridge installations). 2) In installations using piezo cable it’s especially important to make sure all wiring is well shielded, to provide the best available signal-to-noise ratio.

In spite of the lesser signal strength, plenty of customers have gotten back to me to report very successful applications with the cable.

Fuller Tone for Wind Chimes, Tubulons & Other Tubular Chimes 

A few years ago Experimental Musical Instruments put out a book on making wind chimes. In that book I included a photo and a few words about a set of chimes in which the body of air enclosed in tube is tuned to resonate with the chime tone. This gives the chime a fuller and louder sound, particularly in low-pitched chimes. There’s a recording of a resonated chime set included on the audio CD that accompanies the book. I didn’t include a full description of how to make such a thing though; I was trying to keep the book simple and accessible, and I was afraid that a description of the air-tuning process would be too long and complicated. Since the book came out, a couple of readers have gotten in touch with me to say they loved the sound of that chime on the CD, asking why I didn’t give a good explanation for how to make it.

In response I’ve now written an article explaining the process. You can find it here. 

A few months ago I posted information on how to make a pop gun. In that post I mentioned the closely related sound of suction pops, as in the sound of a cork pulled from a bottle. Pop gun pops and suction pops both are very cool and quirky sounds, but suction pops have one advantage: unlike with pop gun pops, the pitch is predictable. If you wish, you can make suction poppers tuned to particular notes. 

In this post I’ll describe how to make suction poppers. Suction poppers and pop guns are quite similar in design, and you can use most of the same components to make either. If you happened to read the earlier pop-gun post, and especially if you made a pop gun, then most of the suction popper making procedure will be familiar to you.

Here’s a sketch showing the essentials of a suction popper.

 To make the suction popper, start with the main tube.  Plastic tubing, such as PVC conduit, is suitable, and it’s widely available and inexpensive at hardware stores. Other sorts of tubing can work as well.  ¾” diameter or thereabouts is good, if only because it’s easy to find corks to fit. The length of this tube determines the popping pitch.  I have made successful suction poppers in lengths ranging from about 10 inches for high pitches to a several feet for low pitches.*

The two corks should have the truncated cone shape, as shown. The narrow end should be narrow enough to go inside in the tube; the wide end too wide to fit. This type of cork can often be found in hobby or crafts stores, if you don’t have any already around. 

The plunger cork must be cut to fit the inside of the tube snugly.  To get a perfect fit, shove the cork into the tube as far as it can go without forcing it.  Cut off the end that’s still sticking out, and sand the cut end flush with the tube end.  This gives you your fitted cork; pop it back out by pushing from the inside with a dowel coming through the tube from the opposite end.

For the plunger stick, you can use any sort of dowel or stick fits inside the tube, about ten inches long. Use a nail or screw to attach the fitted plunger cork to one and as shown in the drawing. Pre-drill the cork to prevent splitting before putting the nail or screw through it, and if the nail or screw doesn’t have a large enough head, include a washer.

That’s it; the suction popper is now ready to play.  Insert the plunger in one end of the tube, place the stopper cork on the opposite end, and pull the plunger back out with a rapid motion. If you’re ambitious, make more suction poppers in graduated lengths to create a scale. Gather a group of friends to play suction pop compositions and improvisations in hocketing style. 

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

 *With longer tube lengths/lower notes, an unexpected effect comes into play: a prominent snap sound appears in the tone. I don’t know what causes this. You can hear the effect quite prominently in the sound sample on this page. The snapping on the low notes comes across almost like a distortion in the recording, but it’s actually part of the natural sound.