William Cumpiano's
String Instrument
Newsletter #13

Go to newsletter number:

Questions, anyone? II

William R. Cumpiano 1998, All Rights Reserved



I have enjoyed your website every time I visit and have learned so much from you about these wonderful creations of art and craft. I am interested in having a guitar built for me and I have so many questions. Below is an excerpt from [Frank Ford’s] website FRETS.COM that I find thought provoking.

Trying to fathom guitar acoustics from a luthier's description of it is like trying to nail Jello to a wall. Frank Ford is a classical example of this. I understand he is an impeccable builder and guitar technician but apparently, not very good at explaining what it is that he does so that it’s useful or even comprehensible. That is not to say that there ARE other luthiers out there that are good at explaining what they know or what they do in a consistent, comprehensible way. The entire literature of luthierie is couched in enormously vague and ambiguous terminology, most of it to hide the fact that luthiers really DON'T KNOW what it is, specifically, that they are doing that causes the guitars they make to sound like they do. But they don't want you to know that they don't know.

Some familiar, innocent, if obfuscatory explanations (luthier acoustics, I call it) are seen in Frank Ford's webpage

In: http://www.frets.com/FRETSPages/Musician/Guitar/ToneWoods/tonewoods1.html
Frank Ford writes:

"Here's a generalization on which most of us can agree: IF ALL OTHER THINGS ARE EQUAL, that is, if we're talking about guitars of the same size and detail of construction, by the same maker, with the same neck angle, bridge height, top thickness, species of top wood, age, then: A wide grain top will produce stronger bass response: That's because there are fewer stiff grain lines so the top is more flexible. From the flexibility comes a lower natural resonant frequency and more easily produced bass notes.

A narrow grain top will have comparatively stronger treble and more subtle bass"

IF ALL OTHER THINGS ARE EQUAL: Where could you really go to get these mythical guitars that share ALL these attributes? Has he tried to test this assumption? Probably not. Then how can he possibly know what he's deducing is true? In the real world no guitars are EVER EQUAL. But luthier's just love to refer to these mythical hypothetical situations which could never, ever be tested.

Plus, for every top with "fewer stiff grain lines" which is "more flexible" I can dig up from the pile a top with "fewer" grain lines which is stiff as the dickens. Conversely for every "narrow grain top" which is stiff, I can dig up from the pile a "narrow grain top" which has the stiffness of wet cardboard. So his generalization (like all the rest on his page) are really quite useless.

Again, I am not detracting anything from the man's capacity to build beautiful sounding and looking guitars. I'm simply pointing out a common folly of just about every guitarmaker that I've ever met or seen, when asked to explain WHAT IT IS THEY KNOW THAT RESULTS IN GOOD GUITARS. A very, very common affliction, which I catch even myself falling prey to at times.

Further remarks based on Frank Fords webpage article::

The backs and sides of virtually all guitars are made of hardwood. Strange as it may seem, softwoods just don't bend well and are more difficult to form into guitar sides.

There are as many hardwood that don't "bend well" as there are softwoods. In fact, many of the tropical hardwoods that are used on guitars fall into the "don't bend well" category, if you look them up in wood technology textbooks.

And there is the matter of tone, too.

Here we go! Hold on to your hat!

Sides just don't enter into the vibrations.

Let's parse that sentence: What does it mean? That they don't vibrate? That their vibrations are unimportant in the general scheme of things? That they don't contribute to the sound of the guitar? Or that he DOESN"T KNOW what their contribution is? Can you compose a sentence more vaguely? It would be hard. Acoustical tests by William Strong at Bell Labs, where a guitar was completely surrounded by an array of hundreds of small microphones, and the array fed into computers in order to show the sonic energy field surrounding the guitar, shows significant sound pressure emanating from the sides at certain frequencies. So they DO "enter into the vibrations."

They define the body shape and support the top and back.

Well, I guess that's pretty significant. There can be no soundbox without sides.

The back, however, is very significant. If you don't believe me, just sit and play a full chord with the back held tightly against your stomach, and then with the back held completely away from your body. Hear the difference in the fullness of tone?"

Okay, so what are we to make of that? Too bad the sentence—and the discussion--ends there. It ends there because he doesn't know what to make of that. But you're supposed to come away with this vague "proof" that the back is important to the fullness of the tone. And with the vague feeling that HE knows what to make of that. Oh, and "fullness" of the tone. Is that anything like loudness? So maybe if I wrap the whole guitar in a wool blanket and squeeze it tightly against my chest, and it sounds less "full" what have I just proven?

Maybe I’ve come down a bit hard on Frank, but I’ve heare luthiers talk acoustics for almost thirty years, and it's invariably done so blithely, carelessly and shamelessly that I just had to comment here. It leads to student and customer obfuscation. And questions like the following one, where I’m asked to comment specifically to Mr. Ford’s careless generalities:

My question, if "sides just don't enter into the vibrations" of the guitar and are there to "support the top and back", would not carbon fiber sides serve multiple needs? Unquestioned structural integrity, visual separation and enhancement of the natural woods?

What is "visual separation" ? A different color? and how does that "enhance" the natural woods?

..and assisting the question of dwindling rare wood supplies? And if it is just a question of structural strength, what are the hardest, densest woods, perhaps alternative woods, that could be used that also have a specific visual interest? Perhaps the following?

The Flindersia family,

Mulga (Acacia aneura) for fingerboards and bridges.

Blackwood: (Acacia melanoxylon)

Blackwood is excellent for back and sides sets. Blackwood is light ochre in appearance to dark with reddish brown streaks. Although it grows from Tasmania to >Queensland, we...

And so forth. I admire all the work you went to dig up all this information and then drop it in my lap to ask me to comment. I graciously decline, however, other than to say that Torres used papier mache for the sides and back of a guitar once, and got what was reported to be an "acceptable" guitarlike tone. But there are other considerations. I can only comment intelligently on practical matters: How easy are they to get hold of? How well do they take a finish, bend, etc.? Or glue? As far as graphite, tt would take a $50,000 heated guitar-side mold to create thin, curved compression-mold graphite guitar sides. (It has to be formed at a very precise temperature and pressure). So that makes me simply not interested in pursuing the matter, since commercially-available materials work so admirably well. If ALL the traditional guitar woods become extinct or illegal in my lifetime, that single maple tree in my back yard will supply me with all the back, neck and side woods to make beautiful-sounding guitars for the rest of my lifetime, and perhaps even the next...

Now graphite makes more sense for a soundboard, because of all the ills and evils that soundboards are prone to, precisely because they are a "natural" material under constant stress which must be at a structurally minimum thickness to function as an oscillator AND a structural member besides. Graphite composites are simply highly idealized soundboard material because they can't expand and contract, crack or take a stress-induced set. And you can make a good-sounding guitar with it. But for sides and back? Too expensive, too difficult to work with unless you're Ovation Guitars.



May I ask some questions ?? ( my knowledge of music/physics is limited to a college freshman course, we only spent a couple of days on the violin & pipe organ) 1. Archtop guitars & violin family instruments appear related ( tailpiece, downward pressure on bridge). How come no sound post on arch-top ?

A violinist supplies a continuous stream of energy to the string via the bow (think of it like millions of tiny picking strokes coming from the bow). The guitar receives a tiny fraction of the player's energy in the form of a single pluck that decays, followed by another single pluck, and so forth. A soundpost, although I've never tried it would likely just restrain the top unnecesarily and excessively. Actually, it would be easy to reality test that, wouldn't it? Just trim a wooden dowel and through the soundhole, wedge it gently between the top and back of the guitar, near the bridge. Then listen to the effect. Or non-effect.

2. On steel & nylon string flat-top acoustics string tension appears to pull UP on soundboard.

It doesn't. The bridge twists around it's long axis, pushing DOWN in front and UP behind it.

Why not distribute mass of the bridge equally above AND below the soundboard and string UP thru the bottom (instead of down from the top) ?

I don't see how that would distribute the mass equally, or why you'd want to do that. I thought (on steel strings, anyway, you were stringing UP thru the bottom (the ball ends of the strings are pushing up from the bottom, aren't they?

3. Does a nylon string classical lose tone &/or volume if some of the strings lay across the headstock instead of feeding directly onto the tuning pegs ?

I can't make that out. Are you talking about slotted or slab pegheads?

Does a slotted headstock offer substantive improvements to tone or playabilty on a steel string instrument?

I think it marginally decreases the output of the guitar, because of its reduced mass. A heavier neck and peghead, I believe, encourages more of the string's energy to be dumped onto the bridge.

4. Sometimes if a sermon wanders, I find myself staring at the laminated wood beams(not very big) which support a large heavy tile roof on my church. Could braces be improved by sawing lengthwise on a sagital plane, reversing > one half, and gluing together. Could they not be made stiffer & smaller ? Would adding a very thin layer of carbon fiber between the layers enhance mechanical &/or acoustic performance ?

Some luthiers I know swear by this. I think that thinning the top and stiffening the top braces is very beneficial to the sound. I just find graphite reinforcements too troublesome to use, and my guitars come out just fine without it.



I have a question about acoustic guitar tops. It seems that every guitar I   have ever seen has a top made of two pieces joined at the guitar's center. A friend of mine swears he has seen a Guild acoustic with a one-piece top. So does anyone ever make guitars with one-piece tops, or was the seam just so well done that he couldn't see it? Is a one-piece top too weak, or is it avoided for other reasons?

Chances are excellent that your friend has been fooled by a well-done seam. I can say with reasonable certainty that no one in the industry is, or has ever in the past used single sixteen-inch wide soundboard blanks (the width of your average store-bought flattop). This is true to the extent that I challenge you to find any soundboard tonewood supplier that supplies guitar top sets much wider than 8 or 9 inches wide. I can think of several good reasons to use two 8-inch boards joined together to form a sixteen-inch soundboard blank:

* They are too fragile to stack, store and handle at that width.

* It is easier to select out flaws from smaller pieces than larger ones.

* Visually and architecturally, book-matching  two pieces along a center-line of symmetry imparts a greater visual and structural uniformity to the finished soundboard. It allows the builder to choose to place the stiffer part, or narrower grain lines of the blank towards the middle or towards the edges. A single, wide plank varies randomly in this regard from edge to edge.



I would appreciate any comments or help that you may be able to offer on a problem that I am uncertain about. I built a classical guitar sticking as closely as I could to the details in your fine book. I am very pleased with the results and the instrument to my untutored ear sounds good. The problem lies in the G string. It sounds true to pitch but kind of muffled, with poor sustain. The other strings have great clarity and sustain. It is not the strings; this occurs with all strings of differing types and manufacturers. If I tune up or down a semitone, it is much better and disappears with a tone either way. I can find no loose braces or seams or other obvious causes.

It is possible that I was perhaps a little zealous in thinning the back plate in one area close to the tail block ( about 10 sqare centimetres area, to fix a bit of tearout from a careless plane stroke.) but the thin area does not appear to have altered the palpable stiffness of the plate and although the plate is still within the recommended limits, as prescribed, I suspect that this may be the cause.( I did not thin the rest of the plate to maintain symmetry as I had a good taptone, and also I did not appreciate the implications of a thinner area.) Could this be it? Can I fix it ? ( I can always tune down, but that is ducking the issue.)

My first knee-jerk reaction is to suspect an overly-thick G string. I would see how a G from an expensive set like a Savarez (maybe try a wound G) or La Bella Professional series sound. Use a lower tension G (i.e., thinner cross-section. All the ones you tried may have been all the same thickness).

Check the pitches with a tuner, all the way up the neck. Chances are that it starts to go sour way before you reach the twelfth fret. If you've skimped on the compensation, the 5 other strings will tolerate it, but the lone G will go way off, and give the "illusion" that it sounds worse than the others when it actually it just sounds sour because of its intonation.

There is no way that I can ascertain why a given string is muffled, even if your guitar was right in front of me. The guitar's secrets reside invisibly within its entire anatomy. But on the other hand, the guitar obviously can't discriminate from all notes just those on a given string, and decide that "I'm going to make one string sound bad". If one string sounds bad, it's logically hard to blame the guitar first. The G string is

always the hard one to get to sound clear on a beginner's guitar because it is slacker AND thicker (thus less flexible and quicker to decay) than the rest, and it calls to the max on a builder's advanced skills and intuition to bring it up with the rest. All the G's on my earlier guitars were dogs. It just got better by guitar 10 or 15 (hopefully with my advice, you won't have to wait as long).

My experience with student guitars (where the students really still don't have a concept of how resilient the top should be) is that because of "learner's fear" theirs top are often overbuilt and overbraced. If so, it might tend to ignore the signals from the slackest string, the G, because it's just not pulling as hard. Thus a few pointers might help (realizing that words are only shadows of reality).

The stiffness of the top is the sum of the stiffness of the soundboard wood, the thickness of the soundboard wood, the stiffness of the brace wood, the height and contour of the braces AND the number of braces. My preference is to choose the soundboard as stiff as you can get it, and then thin it a lot. For very very stiff spruce, .080 isn't too thin. With cedar, I leave it about 15% thicker than spruce. We're trying to reduce that pesky mass, with all that accompanying inertia that the string has to overcome.

I use real stiff brace stock with fewer rather than more braces which remain at their full height in the immediate bridge region and taper quickly to nothing at the ends, a bridge pad, and a real stiff back with at least four largish braces. The top should noticeably "give" on the guitar when depressed with the flat of your thumb just below the bridge.

Good luck!