Principles of Guitar Dynamics and Design

by Ervin Somogyi

I’ve been building guitars for 22 years. I started out building flamenco guitars, which are my first real love. I still play them. Flamenco sank its harpoon into me early on and hasn’t let go; it’s a wonderful music.

The flamenco crowd, I soon learned, is not able to support a luthier. They have enough money to buy themselves Gaulois cigarettes, but that’s about it. The next logical step was to make classic guitars. I quickly found that the classical players are, to the luthier, not a very user-friendly group. They are picky and critical, and since I basically didn’t know what I was doing I found it impossible to please them. It was not a happy experience.

Steel string guitar people have a very different mind set. They are by and large prone to being uncritically friendly, accepting, and encouraging. Their comments are liberally peppered with statements like “Wow!” and “That’s great!” I found their company very appealing. I was able to thus delude myself that I was doing something worthwhile — until 1977, when I was invited to participate as one of seven exhibiting luthiers at the important Carmel Classic Guitar Festival. It was a turning point for me. Seeing my instruments next to those of serious and competent luthiers forced me to reevaluate the quality of the work I had been doing. Up to that point I had managed to live a fantasy and make a very meager living at it. I’ve solved one of those problems since, but I’m still working on the second one.

In spite of their critical stance, the classical crowd has a very useful tool: a disciplined approach which is in large measure missing among steel string luthiers. The classical guitar people that I meet really seem to pay attention to what they’re doing in an organized way. They measure, they listen, they interact much more freely and much more sophisticatedly with guitar players. They have a greater vocabulary in common about tone color, what the guitars do, what they don’t do. Steel string guitar people do not yet have the tradition of this kind of discipline, but I think that will change when the more freeflowing character of the steel string guitar world recognizes the benefit and advantage of it.

The guitar is a relative newcomer to the stringed instrument scene. Before the guitar, the lute was absolutely the most popular plucked string instrument in the Western world. The lute served a very specific musical purpose, which it was no longer able to fulfill as the nature of musical tastes and entertainment changed with the rise of the European middle class. The lute became less and less a courtly chamber instrument and was more often expected to perform for the new bourgeoisie in larger halls and rooms. That was when the problems began. Lutes were really not loud enough. Instrument makers kept making lutes bigger and adding strings, but there were many problems with these approaches. The piano came into being in this period for the same reason: its sound could carry.

As far as I’m concerned, the guitar replaced the lute as the most popular plucked string instrument because it was able to solve the problem that the lute could not: being heard in a large room. It did so through the development of three specific design features: the bridge with a saddle positioned in the middle of the soundboard; the vibrating back; and longitudinal bracing.

Let’s talk about the bridge first. The lute’s bridge is simply a tie block. The forces acting on the bridge of the lute are almost exclusively in line with the pull of the strings. This drives the face in a specific and limited way, tending toward high-frequency, low-amplitude motions. (See Figure 1) The significant thing about the design of the contemporary guitar bridge is that not only is there a longitudinal force acting, as in the lute, but by virtue of the break angle of the strings as they pass over the saddle a vector force is created which actually pushes downward on the face. This drives the face in a different way, creating different tonal possibilities. (See Figure 2)

The fact that the guitar bridge is in the middle of the lower bout creates yet more tonal possibilities, but I’ll get back to this point later.

The saddle, in addition to helping to create the strings’ downward push, is an important coupling mechanism between the strings and the guitar face. If you do nothing to the guitar except change the saddle from a tight fit to a loose fit, you will absolutely lose volume. The fit is critical. Physicists have found that the guitar is a very inefficient sound-producing system. On the average, for each stroke of the strings, about 95% of the kinetic energy thus generated goes to mechanical vibration and is also dissipated as heat and friction. Only about 5% of the moving energy of the strings becomes sound — and if you lose some of this energy at the saddle you will get a disproportionate loss to the instrument. I am concerned that the material the saddle is made of be noncompressible, hard and stiff. I don’t use plastic because I feel it will absorb and damp some of the string vibrational energy. I’ve almost always used bone, except at one point where I was using melamine from cut-up dinner plates.

The average saddle is 3/32″ to 1/8″ wide. I use saddles that are 3/16″ or more wide because it makes a saddle wide enough to offer adequate intonation compensation on a steel string guitar. The rationale for intonation compensation is the subject of another talk, but I want to say one thing about the contact characteristics between the strings and the saddle. If you have two virtually identical guitars one of which has the strings resting on a single high point of the saddle, and the other with the strings resting on a significant portion of the saddle’s top, I think you’ll find this second guitar works better. There is a better and more efficient coupling of kinetic string energy to the saddle by virtue of that extended contact. I can’t prove this, but I suspect if you pay attention to it you’ll get better results. (See Figure 3)

The second feature of the guitar that made it more successful than the lute was the vibrating back. The lute has a sound chamber with its own natural resonant air frequency, as does the guitar. But by virtue of its construction the lute’s sides and back are one piece and very rigid. If you have a second vibrating diaphragm, which the guitar back is, more possibilities come into play because the guitar’s back is active — it actually does something. Try this: put your guitar in its case and gently tap on or near the bridge and listen to the sound you get. This is the sound of the top only, with the back damped. Then lift the guitar partially out of the case so it’s not lying on its back and tap it the same way again. You’ll hear a very different sound. This is the sound of the top interacting actively with the back. In a similar way, when you are playing a guitar the strings excite the bridge and the face and then the air mass, which in turn excites the back, and the back starts to vibrate in some frequency relationship to the movement of the top. The most successful guitars have the back tuned in relationship to the top so that they act together and make a guitar that really projects. In physics this phenomenon is called constructive interference. When the top and the back are mismatched in vibrational activity they are in effect fighting each other, and this is called destructive interference. It is the same phenomenon you observed in high school physics class when you hung a weight from a rubber band and tried to move the weight by tugging on the rubber band: you had to do it at a certain frequency, and when you found that frequency the weight moved with little effort on your part. Otherwise the mass of the weight created resistance to its movement. In this example you are like the guitar top — the driver — and the weight is the guitar back. In the guitar the back contributes to projection and sustain — or relative lack of them, if the top and back plates are working against each other.

The third factor in the success of the guitar was the pioneering use of longitudinal bracing, in tandem with a more centrally located bridge. Lutes had only ladder bracing, with the bridge at 1/6 the length of the face from its bottom. Longitudinal bracing allows the bridge to couple to a larger portion of the soundboard than otherwise and impel it to movement. Positioning the bridge nearer the middle of the soundboard helps in this, because a central point of initial impetus usually means you can drive a larger plate and larger air mass. This simply translates to greater volume. Tap at the edge of any guitar soundboard, and then tap in the middle and listen to only the loudness of response, regardless of tone: you’ll immediately hear what I mean.

The importance of longitudinal bracing is central to the success of the guitar because the guitar is basically an air pump, and in lutherie we need to concern ourselves with how efficiently the guitar can pump air. All other considerations, such as choice of woods and how pretty they are, must be subordinated to this if you want to make a successful guitar.

One way you can get a reading of this air-pumping function is to gently tap a guitar top at the bridge while holding your other hand in front of the soundhole. You will feel a displacement of air from within the sound cavity. You can feel the guitar breathe on you. On better guitars you can feel more air coming out because the top is more responsive; it responds more to the energy of your finger. Dead-sounding guitars won’t be found to breathe on you as much.

A second way of getting a reading as to how freely the guitar is able to pump air is in the sound of the tap tone you just delivered, if you did this exercise. Again, this is very subjective, but still a very useful comparative way to get some sense, some keying in as to what is going on. If the guitar top is tight, stiff, overbuilt, heavy past certain limits (which most commercial guitars are), it’ll sound somewhat like a table top — high, tight, and solid. If the guitar top is free to move, the pitch of the tap tone will go down and the sound emitted will be markedly more open. Obviously there are limits in making soundboards light past which you shouldn’t go because you’ll destroy the integrity of the guitar, but most guitars are so overbuilt that I don’t think you have to be afraid of going in this direction. If you will go around the guitars in the exhibition hall, I’ll be very surprised if you don’t find that the better ones are just more responsive when you do something as simple as tapping.

There are yet other ways of keying in as to how sensitive the guitar is, how sensitive to the vibration and energy of the strings the guitar top and back can be. One fun way is to use a so-called super ball. If something has high internal damping then energy put into the system — in this case the kinetic energy generated via gravity by dropping this ball to the ground — is dissipated by mechanical distortion into heat and internal friction, and this ball won’t bounce back up very high. This particular ball bounces up off the floor almost to its starting point, so it does not have high internal damping. You can get one of these balls and go around the exhibition hall… (laughter) …and bounce this off the tops of some guitars. Some guitars will produce a higher bounce than others. This is kind of a childishly dumb thing to do but it’s a lot of fun (laughter). Some guitars are instructive and you can really learn something from this: the higher the bounce, the more solidly the top is constructed; the less the bounce, the more yielding and responsive the top is. This is really not much different from what the strings do, except that strings are expected to make music when they move the soundboard.

While I believe the degree to which a guitar is successful is in direct relation to the extent that you can free the soundboard up to pump air, this is only part of the story. There are many specific ways in which the guitar top moves, vibrates and flexes in its use of string energy so as to generate a wide spectrum of tones. The lower bout, the area surrounding the bridge, is the main arena for this activity. Let’s examine some of what goes on here when the bridge moves.

How does the bridge work? In what way will the bridge couple with the face? In what way does the bridge transmit the kinetic energy of the strings to the rest of the system so we can ultimately hear music? The guitar bridge moves in three modes: (1) it pumps up and down as a unit; (2) it rocks backward and forward in line with the pull of the strings; and (3) it seesaws sideways around the center line, at right angles to the strings. (See Figure 4) You get a hands-on sense of how much the bridge moves laterally in the following way. Put your fingertips very lightly on one end of the bridge, then tap on the other end. Unless the guitar is heavily overbraced you’ll feel lateral displacement as the bridge rocks from side to side. You’ll feel the motion on one end of the bridge as you tap the opposite end. You may feel a lot or you may feel only a little, but you will feel something. Classic guitar bridges move laterally a lot more than steel string guitar bridges do, by virtue of both construction and bracing.

I’m going to bypass entirely the subject of top selection so let’s assume for the time being that we’ve picked a good piece of wood for the top. Then we have to figure out how to work and shape it. I’m going to try to simplify for you my image of how the top vibrates. I have the innards of a little music box here. This is dime-store stuff but it’s a useful indicator of a fundamental principle of guitar dynamics. (Winds up music box and holds up to the audience). Now this thing here, nobody can hear it. You’re not supposed to hear it. There’s no resonator on it. It’s not exciting enough air for anyone to hear it. The instant you put this on a surface that can take its vibrational energy and excite more air… well, let’s see what happens. (Places the music box on the surface of a table) Maybe you people in the front rows can hear this? Now this isn’t a guitar, it’s a piece of fine furniture. (Laughter). When I place the music box on the top of this guitar the sound can probably be heard a few more rows in the back. So, the guitar will take the energy and excite yet more air with it. And what of the specific sound quality? To make a long story short, what I hear when I put this gizmo on the perimeter of a guitar top is a bright and tinny sound: the high notes are really kind of shrill and piercing. The bass notes in the song are not very distinguishable in quality from the high notes. Now I put the music box, this sophisticated frequency generator, on this guitar near the bridge — a very important place. I notice that there is much more range. Placed over the bridge I can hear lows as well as the highs. This is not a $4000 signal generator, but it shows what a guitar will do when stimulated in different places. If we did this enough we would conclude that driving of the midportion of the guitar face results in a generally fundamental, bassy, full, and loud sound. Activation of the perimeter results in treble activity. This makes perfect sense, as the center of the soundboard is the most yielding part and is thus able to support high amplitude, low frequency activity. The edge, being more rigid, is the logical place for high frequency, low amplitude activity to occur. The top can move as in Figure 5, acting more or less as a unit. That’s bass. The top can also move as in Figure 6, shaking and quivering like a bowl of jello left on your washing machine during the spin cycle. That’s treble.

This is an important clue to brace design. In planning a bracing system, these are the kinds of top motion which you have to plan for, and plan for with respect to a point of primary excitation of the soundboard, in addition to considerations of selection of woods, thicknessing, and bracing. That is, whatever else you do, you have to plan it with regard to where the bridge will be.

Historically, the bridge was placed smack-dab in the middle of the lower bout of early guitars, with a twelve-fret neck sticking out of the body. (See Figure 7) In the 1930s, in response to musicians wanting a more fully accessible fingerboard, steel string guitar makers discovered that players could have an extra two frets worth of fingerboard by making a fourteen-fret-to-the-body neck. This involved shifting the point of bridge attachment that much nearer the soundhole, but bypassed the need to install a cutaway into the instrument. It was a quick fix, and a successful one. But it also required planning the bracing around the new point of bridge attachment. If you install a bridge at the fourteen-fret neck position on a guitar previously braced for a twelve-fret neck you will have acoustic interference from the bracing, and a less than successful guitar. Bracing work involves positioning, as well as shaping.

Bracing is a complicated and never-ending puzzle. I’m sure that throughout my lifetime, and maybe even my children and grandchildren’s lifetimes, controversy will rage about what is the best bracing system. The fact of the matter is that successful guitars have been built with just about every conceivable bracing system. Superb guitars are built with symmetrical bracing systems. Wonderful guitars are built with asymmetrical systems. I’m led to believe that a bracing system as a recipe approach has little virtue. It’s best when it’s part of a context, an thought-out process. To say only that I use X bracing, or Sitka spruce, is by itself relatively meaningless. I will touch on this again in a few minutes.

Traditionally, guitars were strung with gut and came out of a European tradition of individual people like ourselves working in small shops. They paid a lot of attention to what they were doing, the skills being passed on by the best makers into their families for generations. The tradition of making steel-string guitars, however, is American and has almost always been a factory one. Until recently there have been very few independent luthiers making steel string guitars.

You’ve all seen this kind of guitar top before: this is a Martin guitar top. This is the standard today. It is the pattern, the most common way to construct and brace a steel-string guitar in the world. Almost everyone copies it. This is the Somogyi version of the X braced guitar top. Let me talk to you about how these two differ. Before I get into this, though, I want to say that I’m not picking on the Martin guitar design; I have this Martin top because the Martin people were nice enough to give me one when I asked. Other companies I asked a top from, didn’t.

To understand my thinking about guitar soundboard design I want to talk to you about the strength-to-weight ratios of woods and about basic guitar dynamics. The strength-to-weight ratio is a number that expresses how strong or stiff something is, per unit of mass. It is my opinion that it is important to know this about the woods you use. I place weights on my woods when taken to certain standard sizes and measure the deflection; weigh in grams for a measure of mass. Guitar dynamics are the study of how a guitar top moves.

In his book Understanding Wood Bruce Hoadley says that the load bearing capacity of a piece of wood is reduced by 50% when you reduce its height by 20%. This is a pretty startling statistic. So if you take, as an example, a floor joist that is a 2 x 10, and then take another that is a 2 x 8, the latter has only half the load bearing capacity of the former. This strength-to-height relationship has some bearing on the design of guitar braces.

This is a cross section of your average guitar brace. (See Figure 8) Usually it’ll be quartersawn or close to quartersawn. The possibilities for designing braces are virtually endless, ranging along a continuum from low and flat to high and thin, and complicated by whatever contour of scalloping and high points exist along their lengths. What I want to achieve when I make a guitar is done in part by picking woods that have a favorable strength to weight ratio. I want woods that are strong and lightweight. Therefore one way in which my braces are different from standard lies in my choice of materials. Secondly, I and other luthiers like myself are concerned with maximizing the strength to weight ratios of our soundboards. We want to make the guitar as strong as we can while making it as lightly constructed as we can. We don’t want to reduce the height of braces significantly, because as Mr. Hoadley points out, the cost of this to brace strength is considerable. But we want to reduce the mass. I would like to invite you to later step up here and pick up both of these sample tops and compare just how much wood is in each one. You’ll notice that one feels noticeably heavier than the other. My way to achieve this difference is to leave the height on the brace but to lessen mass by making the brace slimmer. This probably does remove some strength, but I believe that it does so in a favorable relationship to the weight that is lost. My approach is to have thin, high braces throughout my guitars. (See Figure 8)

Fred Dickens: You seem to be using the terms stiffness and strength interchangeably. Is that your intention?

I take my woods down to a standard thickness which is greater than the final thickness, put a weight on it and measure how much deflection there is. I call this stiffness and do use this word interchangeably with strength.

About guitar dynamics: are some fundamental differences between steel string and classic guitars, which affect bracing. In part because the steel string guitar is driven by a much heavier metal string, the tendency for it is to be very bright and trebly. The nylon string guitar, on the other hand, wants to be bass-heavy (within the potentials of nylon string response) when left to its own devices. The job of the luthier is to work the wood so as to shift the response spectrum in the desired direction. It’s very hard to make a well balanced classic guitar that has a clear, ringing treble: that’s the whole trick. Likewise, it’s really difficult to make a steel string guitar that has a rich, deep, satisfying bass. You have to do specific things to the soundboards to achieve these things. Furthermore, a luthier has to figure out how to achieve a balanced sound — how to get energy input from the point of primary excitation of the face — the bridge — in both types of guitars to adequately effect both high frequency, low amplitude vibrations and low frequency, high amplitude movement. While there are endless things to be said about bracing systems, plans, and distribution, as well as the selection of bracing woods, I want to remind you of the indicator that we got earlier from that little music box where we heard bright, shrill notes from the edge of the soundboard and heard mellower notes from the soundboard’s center. It reveals that one should pay attention to the perimeter if one wishes to manipulate the high end frequency response of the guitar, and pay attention to the area near and around the bridge if one wishes to manipulate the low end. (See Figure 9) For these reasons you will notice that the profiles of the braces in my guitar top differ from the profiles of the braces in the Martin guitar top, even though their layout is about the same in both cases: my X is lower and my perimeter bracing is higher.

In the steel string guitar, an X brace system is almost always notched together at the intersection of the two main braces. This is an important load-bearing point, and it takes us back to our thinking about strength-to-weight ratios and Bruce Hoadlely’s formula. The main effect of notching the legs of the X brace so they can be fitted together is to turn two large, strong braces into the equivalent of two much smaller and weaker — but still massive — braces. (See Figure 10) This circumstance is bound to affect the response you get from a guitar. The typical factory way of capping the otherwise open X notch is to glue a little piece of muslin cloth on top. If you are concerned with strength to weight ratios and the load bearing capacity of your braces, I think it makes no sense to shoot yourself in the foot by cutting a huge notch in your braces, after all the work you’ve done, without trying to reestablish their original strength by bridging the notch cut. (See Figure 11) Bridging, or capping, the X notch will tie one part of the interrupted brace to the other part, and will prevent the opening and closing of the notch under soundboard movement.

While the X brace is ostensibly designed to have a certain strength, but the minute you cut a notch into it and leave the notch uncapped it’s much, much weaker than before.

One way that I know this is important is by testing it. I made a guitar some years ago which had a capped X brace, but its sound was not satisfying to me. I thought that the bracing was too stiff for the sound that I wanted and I concluded that I would get better results if I shaved the X brace down. I reached in through the soundhole with a palm plane and shaved about 3/16″ off, right through the cap. With the newly-lowered X braces and opened X joint, the character of that guitar’s bass response changed radically. I had changed the strength to weight ratio of the X too much. I recapped the X brace at its intersection and again the sound changed noticeably, for the better. What an astonishing difference removing and regluing a fraction of an ounce of wood made, so long as it was in the right spot. Cutting through that little cap turned out to be a useful mistake for me to have make.

There are a lot of people now capping their X brace joints, including the Japanese. They, by the way, seem to pay more attention to details in the construction of their guitars, even the factory ones, than any American manufacturer I’m aware of.

Audience: have you ever taken the cloth patch from the intersection of the X brace and replaced it with a wood bridge?

I have, but you have to cut back quite a bit to get a flat enough gluing surface to get a cap onto. It’s more of an operation than you might think. One thing you’ll notice on my guitar top is that even though my X braces are tapered in cross-section they are rectangular in cross-section at their intersection: this makes a really tight and very efficient joint. (See Figure 12) I don’t take any wood from inside the notch, as happens when braces are rounded before they are notched together.

Audience: you talked about the shape of the braces but could you talk a bit on the angle of the intersection of the X brace?

My way of thinking is to concentrate on the bridge activity in relation to brace position. These always work in relationship to one another. Let’s take the Martin X brace pattern for an example: the wings of the bridge are normally coupled onto the legs of the X brace as you can see here in Figure 13. Do you recall my example some minutes ago of tapping on one wing of a bridge and sensing the movement of the other wing? In all guitars this is an important bridge motion. In the steel string guitar this coupling is defined by the angle of the X and allows certain degree of bridge rocking. This motion will support the creation of a steel string guitar sound. This sound, then, is made as a function of the effectiveness of the bridge-brace coupling.

In the classic guitar the bridge has more freedom to move in this rocking mode, because the classical guitar usually has longitudinal braces which more or less allow this lateral movement. Those of you who were at Robert Ruck’s workshop might have heard him say that some of his models that had angled fan braces responded differently. It is my belief that these angled braces are dynamically and acoustically replicating the anchoring work of the X brace, which acts to inhibit bridge rocking compared with longitudinal bracing.

Therefore, if we change the angle of the X on a steel string guitar so that the bridge is not so fully held back in its rocking movement, if the bridge is freed or otherwise encouraged to move more actively in this way, then the response spectrum of the guitar will be shifted toward that of a classic guitar — you will get a higher, brighter sound. So that’s my way of answering that.

Colin Kaminski: could you comment on bridge patch size, selection and stiffness and what kind of tone you can expect from these variables?

I had a very instructive disaster some years ago. had built a guitar that had the most wonderful, heavenly bass of anything I had ever made. The treble was OK but the bass had a gorgeous, lovely tone like honey, full and rich, and unencumbered by harmonics that you could distinguish. I loaned it to a friend who mistreated it. He brought it back after a month, during which time he had had another luthier cut the bridge down. This was unforgivably stupid of him and I was very upset. But mainly, the guitar no longer had the tone. I replaced the bridge. I weigh everything: I think one of the most useful lutherie tools is a triple beam balance. Working with strength-to-weight ratios means needing to know how much things weigh, and I don’t want my bridges to be heavier or lighter than a certain optimal range of about 35 grams. Anyway, this other luthier had messed up my bridge so I put a new one on, reestablishing an original parameter of the instrument. But the tone didn’t improve. I didn’t know what else to do so I hung the guitar on the wall. One day I was walking by it and noticed that the light from the windows reflected in the lacquer of the face revealed a noticeable deformation. I became aware of it then because the light was focused on it. The face was markedly distorted and dimpled specifically at the bridge and all around it. What had started as a straight, flat piece of wood had become bent all around the bridge by the torquing action of the strings. And this had been allowed to happen because there was virtually nothing under the bridge to guard against this kind of deformation. At that point I started to install bigger bridge patches on my guitars. (See Figure 14)

One of the things I’ve learned to do in lutherie is to think long term. The guitar is not going to be the same down the road. In a year or so woods will have settled, strings will have pulled something into or out of alignment, and this is one of the forces which guitars are subject to. The distortion in this guitar was caused because I had put in a very small bridge patch, and I don’t do that any more. It cost me the most wonderful bass I’ve ever had, but I learned something from it.

Audience: Did you ever know what caused the loss of that great tone?

No, I never know anything for certain. Everything in lutherie is intelligent guesses. My guess was that because I had noticed a sudden, marked deformation of the face around the bridge, a physical change had occurred which affected the tensile strength of the face and affected its movement. My response t it was to replace the bridge because it had been cut down, and I chose not to put a new bridge patch under it later. I don’t believe it would have distorted as it did had it had a larger bridge patch to begin with — and perhaps it would not have had the same tone then, as the mass would have been different. But whatever it had, it would probably have retained its sound. I kept the guitar on the wall and it was sold to someone who was wonderfully happy with it.

This is another factor in lutherie. The instruments you’re going to like, other people might not; the instruments you might feel indifferent to, others may not be able to live without. So that story had a happy ending. Someone got a good guitar and I learned something.

Audience: you say anything about the domestic woods which are being used instead of the traditional imported woods?

Yes. Do you remember my comments on the nature of the constructive interaction between the guitar top and the guitar back, in a better guitar? Well, the shift from the traditional rosewoods into the “alternative” woods will have to come to terms with this. These are not going to be bad woods, but they will do different things and we need to come to grips with that.

Brazilian rosewood is like glass, and it goes “pinnnnnnngggg” when you tap it. It has very little internal damping, which means that when you put energy into it in the form of a tap (or playing on a string) it’s not dissipated away in heat and internal friction. It’s retained in the form of mechanical vibration, excitation of air, and hence sound. The maples and mahoganies and walnuts have much less of that. So as the back begins to move, if the back has more internal damping, you’ll get a different sound, one characterized by shorter duration of tone. To the degree that you can couple the action of the back with the action of the top you’ll have projection, but this relationship needs to be worked out for each wood individually. There’s no reason you can’t have a good non-rosewood guitar. But I find that talking categorically about woods is not very useful because there can be such great variation within a species. maples are wonderful; others I’m not interested in. The same with walnut. Some koas are as dense as rosewood and some are practically so light that if you sneeze the board blows over. Mahogany, likewise, is much more different from board to board than rosewood is. That’s the virtue of rosewood: it’s much more uniform in working properties. I can only say, good luck to you in your quest for the best sound you can make. Good luck to all of us.

(reprinted from American Lutherie, #36, Winter, 1993)

Original article can be found here : Principles of Guitar Dynamics and Design

GUITARS