Wednesday, April 30, 2008

On Solid Maple Soundboards

I recently saw a photograph of a very beautiful Taylor 6-string. Instead of a spruce top, this guitar had a book-end top made of quilted maple. I have always been under the impression that straight grain wood like spruce or cedar was not only structurally a necessity, but also that it gave a superior tone above other choices. Taylor, being the quality brand that it is, would certainly not make such a basic mistake to overlook the importance of these considerations. Does it mean that my understanding has been wrong altogether? And what type of tone can one expect of an instrument with such type of top? (Could it be that Taylor is merely breaking the mold and becoming innovative - like Ovation has done?)

I personally like a crisp and clear tone in a guitar - almost twangy. I have thought of cutting a saddle and a nut out of aluminum. My logic just tells me that it might give me the type of tone and clarity that I am looking for. What would your view be on this?

Many thanks for all the hours of pleasure you have given me through your book - and also your willingness to respond to my questions.

Cape Town
South Africa

. . . . . . . . . . . . . . . . . . . . . . . . .

The guitar is above all other things a cultural artifact: its form is derived in great measure as a result of historic and cultural processes, from a legacy of dedicated woodworking craftsmen emulating the aesthetic preferences of their day--rather than a legacy of acoustical engineers and scientists pondering over the years how to increase its acoustical efficiency. But in modern times, another process has stepped in to dictate its development: corporate marketing.

In the guitar's cultural continuum, the guitar emerged with certain materials being prized over others. These decisions, again, were based not on acoustical or engineering principles, but on far more mundane considerations: what materials were available to builders, what materials bent the best, or were less reactive, or carved the easiest, or finished easiest, and so forth.

Now every element in the guitars form and structure has an impact, be it trivial or be it significant, on its perceived tone when played. Thus a guitar made with a perfectly straight-grained spruce top will indeed sound differently than a guitar made with a solid slab of jumbled-up-grain curly maple. So, I'm willing to bet, placed side by side and played, there will be a perceived difference. My contention is that a guitar with a quilted maple top falls outside the guitar's traditional continuum. Indeed, I'll be willing to bet that the actual decision of putting a quilted maple top on an acoustical guitar was based on a marketing rule called "product differentiation in the marketplace," and not, "how can we make the sound of this guitar louder, richer, more complex?"

Regardless, you might actually like, even prefer, the sound of the Taylor with its solid maple top. So, if you don't mind corporate marketing's impact on the guitar's historic form, then buy it. But if you think the guitar's historic form has special value, then pass it by. When is a rose a rose?

I haven't heard that quilted maple top guitar, but I would be surprised if the sound quality was not inferior in certain regards. During the 1940s, when Martin made solid mahogany tops on their acoustic guitars--during the period that Spruce was at a premium for the war effort--those guitars which I've played sounded sweet but weak. Yet some other folks--mostly inexpert players--valued them highly. So I would think the same dynamic is in force here: the stiffness to weight ratio of split spruce (it's "efficiency") is greater than any other wood, bar none. That translates to a smaller inertial load being placed on the string signal at the bridge, i.e., less damping. The consequence of less "efficient," wood like maple is that only the stronger, more energetic frequency components of the string signal will succeed in driving the denser, higher-damping top, and that many of the lower-energy signals coming from the vibrating strings will stop there and remain bound up in the string--never to be heard. But just dropping out a fraction of the string signal doesn't only change it's timbre, it changes its loudness. When you drop the sliders in an audio EQ unit, the sound source changes in character (maybe even creates an unusual new sound) and so consequently, the loudness is also reduced as you do so. Ergo, the effect is an unusual--perhaps distinctive--softer sound.

Monday, April 28, 2008

Tap Tones Ad Infitum

I just read your book (very excitedly I might add). I was wondering if it would be practical/possible/useful to post a high quality recording of sound boards in various stages of stiffness and 'just right-ness'? Being an engineer by trade, I'd love to do an objective spectral analysis as a sound board gets progressively thinner, but it turns out that the ears are probably the best spectral analyzer you can get. I take great pleasure in making things with my hands and I take great pleasure in using a guitar (at my ability, I really hesitate to say 'play'). I also love the challenge of making well fitting joinery... I am very seriously considering an attempt at making a guitar based on your book.
Of course, I assume that I would really need to mentally commit to making at least 5 or 6 before it starts coming out right, but hey... I've wasted far more money on wood in the past for projects which aren't even close in the challenge and potential reward factors. (I would need to think about what to do with all these extra guitars!!!)

Thanks (and looking forward to a video?)

Thank you for your interesting message and suggestions.

Your "recording-the-tap-tones" suggestion refers to a soundboard graduation scheme ("quenching the fundamental") that I learned from my original teachers over forty years ago and dutifully included in my Guitarmaking book when I and Jon wrote it in 1985. That was twenty three years ago. I no longer follow that graduation scheme. Clearly my priorities and techniques have evolved over twenty three years.

Your suggestion is a consequence of the fact that the text, unfortunately, can lead the reader to expect upon tapping to hear a clear distinction in the top's response before it is "just right" and at the point that it is "just right." Alas, that is not the case. Over time, I discovered that this was a traditional hallucination passed on from teacher to student over the years. We were all listening to phantoms. The underlying assumption is, that for the guitar to sound "just right," the soundboard must be "just right," i.e., that the soundboard is supreme in determining the correct spectrum of the guitar's response. My subsequent experience and studies in instrument acoustics have since shown me that it is not. But that "traditional hallucination" persists, and pervades the ongoing discussion, like noise.

Sorry to disappoint, but I have come to believe that but the difference between the sound of an ordinary guitar and an extraordinary guitar isn't discerned at a "magic moment" that happens when you're thicknessing the soundboard, or the braces for that matter. That notion reduces the guitar to an oversimplification of its dauntingly complex acoustical architecture.

More accurately, the guitar is a three-dimensional array of linked oscillators--springs attached to springs attached to springs, recursively. The loose ends of the springs are then brought back together to form multiple loops. The string is a spring, the saddle is a spring, the bridge is a spring, the sides and the back and its braces are springs; the neck is a spring, as is the fingerboard and the headstock. They are all coupled to each other, each spring with its own distinctive stiffness and length. Oh, and the air enclosed within the body is yet another spring.

When you pluck the first spring--the string--it's movement drives the entire array to jumping and dancing, attached as it is to both ends. Then, the guitar turns around--because it is vibrating--and in turn, drives the vibrating string, itself in a feedback situation. The driver drives the driven, then the driven drives the driver.

When you plane the soundboard down (or reduce the size of a brace) you are essentially decreasing the stiffness of just one of the springs in the array. Even if it was a "crucial" spring that had to be "the right" stiffness in order to optimize the whole...what would be your reference? What end result is being assured? A "beautiful" sound? What is the reference for beauty? There is no reference for beauty.

Besides, the soundboard is only able to resonate a portion of the guitar's entire sound spectrum, due to it's physical size. It is silent when the string couples to it any sound with a wave length of greater than 22 inches (the soundboard's average size). But the guitar's sound spectrum includes sounds with wavelengths of up to and over 16 feet. Indeed, the guitar cannot resonate any string frequency with a wave length of over 40 inches (the entire guitar's average size). So how does it reproduce the rest of the spectrum? Not by radiation, that is, not by sound pressure waves rising from its vibrating surfaces.

It produces sound pressures at those wave lengths by a distinct acoustical process called "point-source," which relies on the existence of a soundbox with elastic walls, a soundhole and a partially trapped volume of air inside. Its a secondary mechanism that produces all the sounds that the guitar is too small to produce by radiation. So fussing about a stroke more or less of a plane or chisel on some part of the guitar is really a very low-priority endeavor, given the scheme of things. It's very very complicated. Control is just an illusion.

In the end, the guitar doesn't create the beauty, the musician does. The best the maker can do is provide a precise, responsive and durable tool that removes as many impediments to the artists to do their art--as possible. An honorable endeavor in and of itself. Now these are objective goals that can be mastered with persistence and an evolved intuitive sense. But you'll drive yourself nuts searching for the magic stroke that will create art. The least I can do to undo my having misled you is to come clean in this way.

Sunday, April 27, 2008

Can the X-brace Be Improved?

Would you agree that the "half-lap" joint on the x-brace is a weak spot, both structurally and sonically? I was thinking of making an X-brace of two pieces of curved wood rather like two letter C's back-to back, they would have a flat spot where they meet for gluing, and possibly pinning together with little hardwood dowels. (I’d like to send you a small diagram) I think this would give the x-brace strength and continuity, have you ever done anything like this?

I think you're asking the wrong person: I'm not an armchair guitar tinkerer. I'm not inclined to "improve" the guitar by imagining what may be flaws in it's traditional design and then imagining ways to remedy it. My approach is optimize the givens, not change them. But as the saying goes,"I'm a vegetarian, but I'm not going to take the meat off your plate." So you're welcome by me to go ahead and try your scheme and see for yourself.

But you make two assumptions: that the current x brace scheme needs more "strength" and that "continuity" is somehow a goal. Yet some vintage x-braced Martins I've seen dating back to the twenties have carried the string tension load for almost ninety years without distortion. And have you considered it possible that one stiff x brace and one weak x brace is precisely what makes a steel string guitar sound the way it does?

On Bending Cocobolo

Hello Master Cumpiano, greetings from Romania.

First, I will tell you, for me classical guitar is a pleasure.
I have a friend who already build 80 classical guitars. He tried now with cocobolo. But we have some problems with cocobolo. At the bending process of sides; a crack appears at one sides. I read at the others guitars makers web pages that cocobolo have a lot of oils and resins, and water can't soft the sides before bending process. If you can advise us about working with cocobolo will be very glad.

Another questions is the fresh cocobolo bend easily? (he retain a lot of water at begin, with time pass the water goes out and the wood became harder and stiffer, and I think he became more difficult to bend). If I wet the sides in warm water is useful or not? A higher temperature from bending machine is useful to bending easily or not? It is true after the drying process is almost complete the cocobolo doesn't absorb water?

I have allready read about gluing with titebond work real good. The ultimate titebond is better for gluing (Titebond Glue III Ultimate 16 oz )?

I use cocobolo frequently and yes, it is more difficult to bend due to it's oil, and it's stiffness. And occasionally it does break. That is always a danger, no one can protect you 100% from that. But the way to reduce the danger is to be consistent:

1- the temperature of the surface of your mold must be 300-325 degrees Fahrenheit (149-163 C). I don't know what temperature your light bulbs are producing. But the surface has to be 300-325 degrees F. At that temperature, the oils will come out on the surface and darken, but they can be scraped off afterwards.

2- the sides must be reduced to a consistent thickness of .080 inches. That's 2.03 mm. You need a measuring caliper instrument and an abrasive planer. And 2.03 mm is the best thickness: 2.28, or .090 inches, is too thick, it will easily break. The problem is what is the thinnest you can reduce the sides, while retaining their integrity? .080 is the thickness that works for me.

3- I would stay away from "fresh" wood, in any case. Bending fresh wood makes sense only for furniture. But when we're dealing with such thin sheets of wood, it might be easier to bend-- but when it ages, it will be all rippled and twisted. But use only well-seasoned cocobolo that is not crazy-grained, but instead straight and uniform in figure and texture.

4- Apply pressure to the wood in the mould very slowly, let the wood heat up as the pressure is applied. Leave it in the mold for 30-60 seconds before shutting off the power or turning off the gas. Don't leave it in the mold with the heat on for much longer, the wood may begin to darken. The oils on the surface will darken but that can be scraped off Let it stay tight in the mold with the heat off overnight before removing it. It will keep its shape longer. If it changes shape a little bit while it waits to be assembled, I use a hot pipe to touch it up back to the template before assembling it.

5- The difference between Titebond Original and Titebond III Ultimate is not important for guitarmakers. Titebond original is excellent and good enough. Titebond III has advantages that don't concern guitarmakers: its waterproof, you can glue it safely when the room is as cold as 11 degrees C, you can keep the joint open longer before closing it with the glue applied. All those are unnecesary for us. You can go with the Original. Unless you want to use the guitar under water!

On Pinless Steel-String Bridges

I am working on designs for a new guitar and would like to try a pinless bridge. I like the simplicity of string changes and the clean look. In your experience does this type of bridge have a marked affect on the tone of the guitar, and also is a bridge plate still necessary?

I'd definitely avoid putting a pinless bridge on a steel string guitar.

Pins anchor the strings to the bridge and soundboard and not to just to the bridge. Ball ends pulling up against the bridge patch and pins hooking into holes in the top result in a more secure mechanical connection between strings and soundboard. Having the strings end at the bridge dumps the entire load 100% onto just the bridge's small glue seam. I've just seen too many Ovation Balladeer bridges peel off their soundboards to try messing with an old, well-working tradition.

Indeed, if I were required to use a pinless bridge, I would doubtlessly use a bridge patch. Few people realize that the bridge patch also functions to stiffen the soundboard under the bridge so it's harder for the soundboard to just peel off the bridge under stress. The bridge patch protects the integrity of the bridge's glue seam besides providing a hard bearing surface for the ball ends.

But it won't stop the glue seam from failing because too much load is being directly dumped onto too small a gluing area. The bridge would probably not just peel off, it very well may pop off. But if you must use a pinless bridge, increase the size of the glueing area by increasing the size of the bridge over that of the standard pinned bridge AND make sure to include a sizeable bridge patch.