William Cumpiano's
String Instrument
Newsletter #22

I have been building a guitar over the past several months using your book as my guide. Everything seemed to be going very well, but when i left the top plate sitting in my garage on my workbench for about a week or so it warped. The top was completely finished with rosette and all bracing in place and shaped. If I hold a straight edge across both the lower and upper bout i have a gap in the center of about 3/16 inch. It seems as if the top plate has shrunk and force the bracing to bow. Is there a way to fix this or do i need to start over and how do I avoid this in the future.

Sorry you have a problem.

You must have skipped page 157, the discussion titled "humidity". The reason for your current dilemma is stated plainly there.

Without a climate-controlled room (air conditioner, room humidifier, room dehumidifier, all set to 70 degrees and 40-45% humidity all the time), you are at nature’s fickle mercy. I'm afraid there is no way of undoing nature's work--other than:

1- Carefully chisel off all the major curved braces and upper transversal face graft.

2- Prepare your new top braces but don't glue them on. Proceed to side bending, and fabricate and gather all the soundbox parts before actually assembling them together. You can take your time making them in an uncontrolled humidity environment.

3- Reserve a long weekend, or a time that you can devote a full three or four days full time to the guitar. Oh, and that 3-4 day period CAN'T fall on a rainy period or the hottest time of the year, i.e., during any extreme weather period.

4- Assemble the entire soundbox according to instructions in the book over this collapsed 3-4 day time period. Once the soundbox is built, you shouldn't have to worry about humidity any more.

That's how good guitars are built. I'm sorry it's so much trouble...

Side-bending tips and options
>Do you bend all your sides with a hot pipe? Woods such as maple seem to be very difficult to bend without damaging the wood. What do you think about the side bending machines such as the Fox side bending machine heated with light bulbs? I have heard there is a small amount of spring back, I have never built a guitar using a mold, do you think the spring back would be a problem using the free standing method described in your book. I have been thinking about adding one of these machines to my shop, and I would really appreciate any information you can send me. Thanks.

I have four or five side molds for the more common shapes that I build, and I use a Luthier's Mercantile Side Bending heating blanket (with a rheostat) laid on top of them. I clamp a shaped shoe down on the side slat at the waist and clamp the ends to the mold. It works rather well. I’ve seen others use the Fox light bulb bending machine to good advantage.

The trick is to let the sides "cook" for half a minute before shutting the heat off, and then leaving them to cool and dry thoroughly for two to four hours, if not overnight, before unclamping them--which reduces spring back in most woods (there is virtually no springback if you let them sit clamped overnight to fully cool and dry). I touch up the residual spring-back on the hot pipe just before assembling the sides to the top.

For the unusual, or one-of-a kind shapes, I use my bending iron. For this I avail myself of my many years of experience to get them right with a minimum of damage or scorching. It takes practice...what more can I say? You get a feel for when the wood is getting too hot, or when it is about to break. It’s easier to demonstrate than to describe. More like choreography.

But the mold and blanket, or light bulb system is pretty foolproof and an excellent option to people who feel life is too short to learn hot pipe side bending... and willing to make just two or three different size and shape guitars, only the shapes that you have a mold for.

Here's some useful tips:

* Bending at the right thickness is key here: the difference in bending ease between .080" and 090" is vast.

* Your chances of success improve if you know what the temperature is, and can set it with some accuracy. I have a digital thermometer (pyrometer) and don't start to bend below 250 degrees F. Above 300 degrees darkening begins and at about 325 to 350 scorching begins. Clearly the wood can bear more heat moving over a hot pipe than sitting immobile over a heating blanket. So contact time is a factor.

Maple stains easily on the hot pipe AND on the molds. The main thing here is to clean the pipe of old dark resins from bending other woods and use strictly distilled water. The minerals in tap water boil deep stains into the pale wood. But Maple, as all the temperate climate hardwoods are especially easy to bend actually. It's the tropical hardwoods that are difficult, especially mahogany (it tends to fold at tight bends instead of bend) and some of the extremely dense rosewood hybrid species (which spring back in major ways and you have to bend and rebend three or four times).

Yes, it’s a challenge. If it were easy, you'd probably not want to do it, right?

>When I use my bending iron to bend the tight radius in the middle of each side I get a somewhat splitting almost like a pulling or cracking just about in the middle of the bend. What am I doing wrong? Is the iron to hot or is the wood (mahogany) not thin enough (.008-.009). I am using a pipe that is approx. the same radius.

"Somewhat splitting, almost like a pulling or cracking"... ah, yes, mahogany... If it's creating what looks like a wandering cross-grain split, right at the stress point, it's the tissue collapsing. For any side material to bend, one face has to stretch and the other has to compress. Mahogany can stretch far more than it can compress. In this way, mahogany is in the “bad to bend” category as are most tropical hardwoods (Northern hardwoods, like beech or birch or maple are in the “good to bend” category). The tissue is folding into itself right at the spot where it’s being forced to compress. If you're lucky, the lines can be scraped away if they're not too deep. Otherwise...you'll have to throw it away and start again. Next time change your technique by one or more of the following:

* Change the waist template contour to a less radical bend. Big help. Why do you think Martin loved Dreadnaughts so much?

* Don't soak so long; or don't soak at all and just spritz it generously as you're bending. Overwetted mahogany collapses easier.

* Thin the blank further just in region of the bend. There's a world of difference between .080 and .090. Tight waisted bends are very difficult at .090, far easier at .080. If it’s a deep body guitar, don't go much under .085, or your sides will end up looking like wavy gravy. If you don't have the capacity to gauge the wood that closely, well...develop it or live with the results.

What color cheese is the moon made of?
>Which bridge pad material allows for a brighter tone [on a steel string guitar]? I was doing a web search in the hopes of finding out what different guitar builders and repairmen's opinions were on what wood allows for a brighter sound in a spruce top dreadnought guitar for use in the bridge pad; rosewood or maple.

Your query is phrased not as a question but as an assertion. It asserts that the bridge pad specie determines the brightness or darkness of the tone. So now, which species makes the guitar brighter?

Your question is not exceptional. I get these all the time. They ask me to clarify some detail or other of an assumed, shaky premise. What color cheese, exactly, is the moon made of?

Forgive me: I'm not trying to be a wise-ass. You’re just asking the wrong guy. You should really ask the guy who convinced you that the bridge pad affects the guitar’s “brightness”. No doubt he’ll lead you even further astray.

As for me, it’s a moot point. I’m content to choose species hard enough to keep the ball ends of the strings from driving their way through the soundboard. As for its effect on the “tone”, I can only comment that there is just no way to ascertain how changing any single element on a guitar is going to change it’s sound, when a guitar is composed of hundreds of interacting and interconnected elements—some of which you can consciously make decisions about, and others that you can’t. This particular builder can't tell you--or if he could, can't describe to you--the effect on the whole of changing just one of those decisions.

The sound of a guitar is the sum total of hundreds of decisions that the builder has made—and a series of givens the builder must simply deal with and accept. Some of these “givens” are the traditional form of the guitar. The bridge patch is there, traditionally, not to change the tone, but for more practical reasons. Attributing an after-the-fact precise and predictable tonal function to it is one of those “shaky premises.”

The other problem is in describing sound with words: I really can't tell what sound you're hearing in your head when you ask me about a "brighter" sound; or a "warmer" sound. Or a “darker” sound. Or a "clearer" sound.

Yes, my answer is unsatisfying. I could just tell you that rosewood bridge patches “allow for” a “brighter” sound. That answer would be more satisfactory, no doubt. But then you would be getting your information from a faker or a fool.

Why can’t an acoustic guitar be more like an electric?
>It seems to me that electric guitars have much lower action AND less string tension than acoustic steel string guitars. Is this just my imagination or is it just due to the fact that I have never played an expensive or custom acoustic guitar that achieves low action without fret buzz?

At a given pitch a thicker gauge string produces more acoustic output when plucked than a thinner gauge string. It also feels stiffer because a thicker string is less elastic than a thinner string. Thinner gauge strings are usually put on electrics because you don't need the acoustic output. You just turn up the amp if you want more volume. So electrics feel softer to play. And when you play an acoustic, you have to pluck harder because you want to hear it. On an electric, you don't pluck as hard because you can hear it by just turning up the amp. If a player (acoustic or electric) usually plucks softly, the action can be set lower. If his technique is to play harder, you have to raise the action, or he’ll complain about buzzing. If an electric player plucked his strings as hard as an acoustic player has to, and his action is set very low, it would buzz. You can't expect to set the action super low and then pluck hard on it, and not get any buzzing.

Lower action with less buzzing can be achieved on both electrics and acoustics if the fretwork set-up is expertly done. Again, in either case, you can force a buzz by playing hard enough to drive the string to hit the fret. An expert player can play more vigorously with low action because he/she knows how to strike the string in a way that buzzing is reduced. So technique is a big factor here also. Indeed there are many other factors, such as the effect of the scale length and the resonances in individual guitars that can make a string's vibration increase (the string sets the guitar in motion. But remember: the guitar is the platform for the strings. So it feeds energy back to the string. Acoustic guitars are made to be as compliant to vibration as possible, electric guitars are not. So the feedback effect is exacerbated on acoustics—generally requiring higher action than on electrics). You have to look at the individual case.

And yes, a more skilled makers/technicians can optimize action by bring their experience to bear. They can indeed, lower the threshold at which a string starts buzz at lower action settings on any given guitar by improving the guitar’s fretwork and set up...and instructing the players how their technique can result in cleaner playing.

Is it “allowable” to radius a classic fingerboard?
Thanks for writing your book and providing the informative website. I am a new builder. Classical fingerboards are typically flat (as I have found in my research to date). Is it blasphemous to put a radius on a classical fingerboard? Are professional builders selling classical guitars to students/guitarists with radiused fingerboards?
Is this too great a break from tradition?

Not too great. Classical making has become far less hide-bound than it used to be, and if a slight radius provides greater comfort for the less expert musician, go for it. We actually radiused a board—to a dramatic degree—on a classical belonging to a recitalist that had developed a condition in his barring finger that prevented him from straightening it (“trigger finger” it’s called). We actually curved the board to precisely match the contour of his locked barring finger.

He's back in business, and by the way, it turns out to be fabulously comfortable for non-ailing guitarists who don't practice a lot (like me), although the required arch on the saddle and the resulting curve of the string array takes some getting used to for the right hand.

Keeping the headblock angle fixed is crucial!
I'm a custom furniture maker trying my first guitar and I have question, if you have time to reply. I understand that the curved upper face brace should impart enough dome to the soundboard so the end of the fingerboard will sit correctly when the neck is backset. But what about the headblock? If it's top is flat, then the first two inches of the top will be flat. Should the front of the headblock (and the front of the upper bout) be shaved down slightly? I don't know if affects anything, but I'm building in an outside mold rather than on a workboard.

PS: your book is dangerous. I'm having a hard time working on the paying jobs in my shop since I got it.

Congratulations--you're the first beginner that has successfully visualized that esoteric spatial relationship. It has been missed by something like a thousand inquirers since the book was published in 1985.

Well, what actually happens is that the rise in the top (towards the soundhole) causes the headblock (and attached sides) to rotate minutely. This can happen because there's a bit of flex in the system. And that's okay. The result is that the outer face of the headblock will not end up strictly perpendicular to the workboard, but tipped back by a degree or so. It's okay because the neck itself has to be tipped back by a degree or so, and that should preserve the perpendicularity that is required in the region of headblock, which should approximately match the perpendicularity that exists between the heel bearing surface and the fingerboard tongue.

But problems can occur if you inadvertently increase the headblock angle during the assembly process. That small angle offset should remain after the back is glued on, because its the event which locks the headblock's orientation in place forever. Clearly, if the headblock has been forced back too far, it will force a slight hump in the top where the fingerboard will be glued. If the headblock ends up perpendicular or worse, tipped FORWARD, you will end up with a hollow in the area above the soundhole.

In what ways, then can the headblock be forced to end up in an improper angle? If you are using the workboard shim, you have to be careful when you add the additional shims under the upper transversal face brace that, when you clamp the guitar down with the workboard shoe by tightening down the wingnut (that is drilled through the soundhole), that you do not rack the headblock back. I just start with an oversized shim and sand it thinner progressively until I can snug the wing nut down without watching it pull back on the headblock.

Since I announced in my website that I was recommending a barrel bolt neck attachment system over the pinned mortise and tenon system (were you aware of that? Details are on my website.) I can now take advantage of the large bolt holes that now appear on the back wall of the headblock. They allows the headblock to be screwed and thus fixed securely to a vertical extension to the workboard shoe. through the assembly process. The trick is to trim the shoe extension to the same angle as the headblock tip-back. A little trial and error will get it right. After the back is glued on on, remove the screw that tightens the headblock to the shoe, and after unscrewing the shoe from the workboard, the shoe can be worked out through the soundhole.

Scale length considerations
>Can you tell me where I can get some information on short scale (630 to 640mm) classical guitar construction? I would like to have one built but would like to learn first what should be considered in terms of body size, bridge placement, neck size, string distance etc in order to get an instrument with good tone and volume. I have small hands and would like a guitar that is easier to play but still sounds good.

The information you seek is rather specialized, I doubt that there is much published on specifically short scale guitars. But I can venture the following comments: Shorter scales certainly facilitate long reaches, as would fingerboards that are less wide and necks that are less thick. The last two have little effect on the guitar's performance (much more on the guitarist’s) but there are serious consequences to shortening or lengthening the scale on a guitar; some would say dramatic. Some of the consequences can be mitigated, or balanced by changes in the soundbox design, but not all.

A change in string tension is the greatest overarching consequence. You have to wind a longer string tighter than a shorter string to have them both reach the same pitch. Thus, a shorter scale guitar has its strings somewhat looser than a longer (even when both are tuned to the same open-string notes). The amount of energy the strings are applying to the soundboard is thus less, and not just that: the rise and fall of the plucked note's sound is rounder (as contrasting to the "snap" of a long scale string); and the harmonic content of the shorter string's signal is proportionately richer in higher frequency components--and viceversa. So we may say that long scale guitars have a more "baritone" quality and short scale guitars have a more "tenor" quality to them. So there is variation in the sound character--proportionately--among shorter and longer scale guitars.

There is some difference in opinion about loudness and projection difference. Where one would expect that the greater power inherent in greater tension would yield a larger tone with more projection, in fact a shorter scale guitar typically produces its power in the higher frequencies, in the ranges where the human ear is more sensitive, so many small guitars seem disproportionately loud for their size and their voice seems more penetrating or "piercing" than a deeper-voiced instrument.

The "quality" of the sound (rather than its character or relative loudness), I feel safe in declaring, has little to do with the scale length or neck contour, but more to do with the quality of materials and appropriate design and execution of the soundbox. That of course, depends on the skill and experience of the luthier, who should be able to make a beautiful sounding instrument, from a 12" scale guitarrillo to a 30" scale guitarron, and all shades in between.

Lowering the action by shaving the bridge
>Could you please comment on this alternative method to an acoustic guitar neck reset? I hear quite a few repairmen talk about thinning and recountouring the bridge and ramping the string holes so that the saddle can be lowered even more and proper string angles retained. I think that, while this repair may solve the problem temporarily, a neck reset will be required anyway in a few years time, and then the bridge will have to be replaced, too. Maybe this technique will give some life to cheaper instruments (considering the high cost of a neck reset). What do you think?

There is no doubt that some repairs are appropriate for expensive and valuable instruments that aren't appropriate for less expensive instruments, and vice-versa. But since there are so many factors involved, sometimes there is a difference of opinion about which repairs are appropriate in each specific case. So it is difficult to assert a rule that covers every possible case.

Thinning and re-contouring the bridge is appropriate for inexpensive instruments if it will return it to easy action at a low cost. But even if the instrument is expensive, thinning and re-contouring the bridge may be also appropriate, but only if, in the judgment of an expert luthier, it is excessively massive to start with, and if the process may actually improve the instrument's acoustic response. Thinning and re-contouring is never appropriate for a vintage or historic instrument.

Neck resetting may be economically AND physically impossible on inexpensive instruments. There are even some experts who maintain that neck resets are bad even for good instruments because of the potential stress and risks involved. They maintain that bridge thinning and re-contouring is the least stressful solution, thus the most responsible on a valuable guitar in every case (Are you out there, Matty?). But this opinion is rare, although there can conceivably be certain specific instances where they are indeed correct.

The problem ultimately with thinning and re-contouring is that if the instrument is built in such a way that it's distortion is progressive (which is the case with many, but not all, steel-string instruments) there will inevitably come a time when the bridge cannot be further thinned. A neck reset is then the only solution, in which case, yes, the bridge should be replaced also. On an inexpensive instrument, the most economical solution may be to "put it out to pasture."

Lacquer finish dried soft
>I have finally after a few delays finished my guitar with lacquer as suggested in your book. Two coats sprayed on, wait 24 hours, wet sand, two more coats, wait 48 hours (actually more) then wet sand. The final wet sand was with 600 and then 1200 grit wet paper.After sanding I buffed the surface with an automotive compound designed for lacquer. The finish buffed out beautifully! However, here is my problem. When I lay the guitar down on a surface, such as a cloth, towel, etc (never a hard surface) the cloth will leave light impressions in the finish after a couple of hours or so. The finish feels solid so I am sure it is dry. I did notice that his did not occur after sanding and prior to buffing.

Do you have any advice on this? I am at a loss. At this point my only solution is to try to buff out the light impressions, which does not really work well. It is necessary to re-sand the surface with the 1200 grit paper to totally remove the impressions. These impressions are light (just visible, and barely felt) but they obviously mar the finish. I appreciate any advice you can offer. I used a furniture grade lacquer purchased at a specialty paint store. I did not thin the product. I sprayed it on with an automotive type spray gun.

You never said how long after applying the last coat did you wait before rubbing it out. Or how long it has been since you rubbed it out. Or the brand of the finish.

I waited about 72 hours after the last coat. It is manufactured by a company here in Fort Worth, Texas called Trinity Coatings. It is a professional lacquer. I also failed to mention that I applied grain filler and sanding sealer before applying the finish coats. Also, I rubbed it out by hand.

I would first wait another week or two and do the cloth test again, and if you still get an imprint, you should call the techie at the the company and ask what you did wrong or what the problem with the finish itself was.

My guess is (a) you didn't stir the stuff in the can sufficiently b) the batch of stuff in the can was improperly formulated (too much plasticizer, or too much slow-drying solvent) or c) the stuff was improperly stored in the warehouse.

My experience with this problem is that the finish eventually hardens, even though it may take months. By the way, it is always best in every case to let the instrument sit for one to two weeks (better still, three or four) after finishing and before polishing.

Why is my guitar dead in places?
>I have just completed making an acoustic steel string guitar and a classical guitar using your book GUITARMAKING, TRADITION AND TECHNOLOGY and I have a problem I need advice on. On both guitars, on the first (E) and second (B) strings, the sustain and overtones begin to drop off from approximately the 8-10th fret up. By the time you get to the 14th/15th frets, these strings are very dead. Can you tell me what the problem is - - can I correct it on these two guitars and can I do something to avoid it happening on the next guitar?????

Your question is pretty impossible to answer as asked: the guitar is way too complicated for me to be able to diagnose a problem by email with the description "the overtones drop off" or the "strings are very dead." Having said that, I will venture that your frets or fingerboard may not be securely seated in that area.

Can this Yamaha be saved?
>I don't know how you feel about older Yamaha guitars, but my father has an FG-512 12-string, and 6years ago(the last time I played it), it sounded great and played well. Fast forward 6 years, and it has incredibly high action, and a low bridge height. I adjusted the truss rod as much as I dare, but it only helped slightly. I suppose the neck angle has changed. This guitar has been sitting in a mini storage locker for about 2 years residing in a chipboard case, and is about 20 years old. My question is this: What can I do to try and rectify the problem on my own, do I need professional help ,and is there hope for this guitar? It has a really nice sound and is a nice looking instrument, so I hate to get rid of it. I think it uses a plastic glue or resin glue, so it is not easy to do work on.

I'm an old hand with those particular guitars. It's what they all do. I'm afraid you have to have fairly advanced technical skills to pull off a neck reset, which involves loosening the plastic binding-- without breaking it-- from the seam near the heel, in order to reveal the back seam around the entire front of the soundbox; then sawing through the back seam [as well as the back/headblock seam] in that area with a razor saw. This frees the neck, allowing you to pull it back in the direction opposite to what the string tension pulled it to, and then glue the seam shut while the new (and appropriate) angle is accurately held in place.

The exposed edge of the back plate will now project about 1/16" into what was the binding slot. It then has to be trimmed away so the binding can sit back into place. The binding is glued back and the chipped finish touched up. The instrument emerges as good or better than new, with only slight evidence of the surgery. The technique is called "shortening the back" because you've now you’ve actually shortened the back while tucking the neck, neckblock and sides into the soundbox.

We haven't done one of those for a dozen years, they just don't seem to show up any more. But that's what's involved. We could offer to do it for you, but it's likely to cost about what the guitar is actually worth or somewhat more. Even before shipping and handling fees.

What do I do now, Uncle Bill??
>I made a stupid mistake last night gluing up a back on a classical. I put too much pressure with a clamp on one side- I used the clamp to augment the bands I was using- never again. Anyway the pressure popped the center back brace loose just at the point where I arched it. The rest of the brace is intact and it did not affect the arch. However, the brace is slightly raised at that point near the side. It did not damage the kerfing as the brace just touched the kerfing and was not under it. So- my question is, should I try to reglue the brace through the soundhole by using a jig to force it down to the back or could I just fill under the brace with a small wedge and glue to make sure I don't end up with a rattle. Or- the least desirable option, should I take the back off, fix the brace and reglue. This luthier art is a constant learning experience. I spent hours getting the back
ready, re-reading the chapter in your book. And then the failed back brace.

That's a tough one. You shouldn't have to go through the hell of removing the back, not that it wouldn't be good practice if you were ever going to become a repairman some day. …Are you? Well, with that out of the way, let's see how I can guide you, sight unseen, by email.

The middle brace, huh? you can probably see the problem right through the soundhole. That's why you've offered to put a wedge under the brace, because it right there in plain sight, right?

No, you don't want to but a wedge between the brace and the back. You want to put a chip at the end of the brace so that it's locked in , like the ends of the other braces. So how about this:

1- You want to create a "jack" which is a stick that spans the space between the top of the tip of the brace and the soundboard. Then you want to wedge it between the brace and the soundboard so that it presses the open seam shut (when there's fresh glue in the seam). Make the stick pretty thick, about 3/8" square (to spread the pressure) and round the soundboard end slightly so that it lifts into place without snagging on a sharp corner. You make a brace that is a bit too short, and then keep adding strips of masking tape to its end until it is long enuf to press into place. It should be just long enough to flex the popped tip down against the back.

2- Do a few practice runs until you can educate your hands to slide it in through the sound hole and prop it in place smoothly and securely (you don't want to drop it into the guitar and search for it while the glue is setting up).Now remove the jack.

3- There are several ways I can think of to get glue under the brace. I'm presuming you can see it through the soundhole, and that it is actually pretty close to it. One way is with a thin artists spatula or palette knife, the type with the offset handle, not the straight handle. Or a bent, thin feeler guage, which is what my partner Harry uses. He glops glue on the up-surface only and slides it in there under the brace. A couple of applications should do it.

4- Apply the jack and let dry. You can place a slat directly over the brace, across the top, clamping across the top and back with a padded cam clamp, snugging the back down gently over the prop and glued brace until the glue sets.

5- Try to carve the kerfing chip above it off, and make another chip that it bigger so it spans the distance to the brace end. Glue it in place with white glue holding it firmly against the side and against the brace end with your fingers while counting slowly to sixty (better ninety). That should do it.

Is there an easier way to thin the plates?
>I'm embarking on my first classical guitar and thus far I am enjoying the experience. I find your book to be of tremendous value to my education. Now that I've buttered you up ,I am struggling a bit with thinning the back and sides. In your book your mention that if you are getting chip outs and such that you should take whatever time is required and sharpen and adjust the plane. I've made several attempts at sharpening and even purchased a book specifically on sharpening, I believe I've made progress on sharpening, specifically with the scraper, but I continue to get chip outs and gouges. For the back I resorted to sanding the back with a long block sander and alternatively scraping with a scrapper until all of the gouges have been taken away. This was a long and labor intensive exercise. I'm just now starting the sides and have experienced the same problem with gouges and chip outs. I'm more concern about the sides since I eventually have to bend them. Do you have any words of wisdom that can help me? I don't want to get discouraged.

That’s a fundamental guitarmaking problem that all aspiring luthiers must resolve before going on to the next step in their evolution. Your own personal evolution from a beginner into a fine instrument craftsman that can shave thin slats of hardwood successfully is analogous to the larger process of how instrumentmaking evolved from a rustic folk craft that dug out a solid block of wood to make a soundbox—into a refined craft-guild profession that learned how to razor-sharpen hand planes to easily shave thin slats of hard, figured wood that could then be heat-bent into elegantly curved soundbox plates. Ontogeny recapitulates philogeny (look it up).

You didn't say how thick your blanks were to start with, but there really is no other hand technique other than planing or scraping. Donald Brosnac actually placed a hand-held belt sander on his plates, and had no compunction to teach others to do the same—but that seems like a recipe for disaster. Clearly scraping is safer than planing with a poorly sharpened and set-up hand plane. Indeed, there is a tool called a scraper plane that allows you to hog greater amounts of material, but that also requires some set up and sharpening know-how.

The problems you are having are shared by most luthiers and is why most of them (the ones who are too lazy or busy to develop their plane techniques) eventually obtain some sort of abrasive planer, either home-made or store bought. The Guild of American Luthiers offers information about shop-made devices that different luthiers have devised. These devices are generically called "filletiers", from the French. The other alternative is to check around in cabinet shops in your area and ask whether any one has an abrasive planer (or a wide-belt sander) can reduce your plates accurately for a fee.

Fretboard surface geometry: ruminations
>I have a question for you concerning the fretboard surface geometry. In my day job I program and run CNC milling machines. It is relatively easy for me to develop surface geometry with compound curvature on computer, and to generate toolpath to cut the material on the CNC.

I have read somewhere about curving the fretboard from end to end to account for the waveform of the vibrating strings, and to vary the cross-sectional curvature for a more comfortable feel when fretting the string.

I am generally assuming the curvature could be derived from understanding the waveforms of strings 6 and 1, then allowing the software to interpolate the transitions thru the other strings. I don't know, however, about factors like neck deflection due to string tension, how to measure the string waveform, how much to curve the cross-section if at all, etc.

I am contemplating generating toolpath to apply this curvature to the fretboard and to cut each fret slot. The slots would also conform to the curvature accurately. The mill has very good precision (x.xxxx").To this end, can you recommend any good method of determining the correct curvatures?

Luthiers, at least myself and all the ones I know, approach fretboard geometry from a far simpler vantage point. We let string tension curve the fingerboard to create "relief" (clearance, which will occur naturally as a result of six strings pulling up and slightly curving a slender neck). The curvature is then controlled back with an adjustable truss rod towards a contour that is more appropriate to the player's attack. Heavier attack, more curvature; lighter attack, less curvature. The adjustable rod is necessary because the precise response of a neck, of precise stiffness, to the precise tension, of the precise set of strings, of a precise string length--is pretty inscrutable to old-world woodworkers like us. Thus the great utility of an adjustable truss rod.

For almost a century, The Martin Company postured that their necks responded so consistently and predictably that an adjustable rod was unnecessary. During the nineties they finally gave up that fiction, after for decades dealing with expensive neck warranty repairs that could have otherwise been simple adjustments if they had deigned to use adjustable truss rods (they instructed their warranty repair shops to adjust the neck curvature by refretting part of the neck with frets with larger tangs!!!) Well the new generation has gotten rid of all the old problem stuff that the old would never change because the older was so hidebound. But I digress…

The common visualization of the longitudinal curvature of the fingerboard is that the string generates an "envelope" when it's plucked which looks something like an elongated leaf shape, and the neck's curvature should "mimic" the strings curvature. This allaws the string array to be adjusted closer to the fingerboard without buzzing. If the neck remains perfectly straight after the guitar is tuned up to tension, the action will have to be set very high to avoid buzzing on the lower-pitched notes when plucked energetically. The harder the string is plucked, the wider the leaf shape. As I understand it, the lower the pitch of the string, also, the wider the swing of the string. So the rod is adjusted to accommodate the worse possible scenario: it must give the low e sufficient "room" to vibrate when its plucked unfretted. As you fret it closer to the body, the fingerboard surface will tend to go toward straight, which is all right because at a higher pitch, the vibrating string doesn't require as much relief. So the idea is to tighten the rod (straighten the neck) until the player, playing energetically (each player has a different idea of what "energetic" is) starts to hear a rattle, then backing off a bit. That is what is meant by 'the fretboard should follow the shape of the vibrating string'. But the shape is not cut into the wood. It's generated by the string tension and then controlled by the adjustable rod.

By the way, there is another intractable variable that many builders account for: the neck and fingerboard more or less consistently being bent BACK by the wedging action of the frets. Some luthiers account for this by gluing a longitudinally straight and already transversally arched (and slotted) fingerboard to the neck; then tightening the rod to induce a BACK bend into the neck; then hand planing the hump off along the run of the strings until the fingerboard is straight (but still arched) again. This "loads" the neck and gives the adjustable rod action in both directions. Unfortunately, this is not something that can be CNC'd very easily.

So in the end, I shape the fingerboard straight along it's full length, along the run of the strings. String tension and the truss rod will take care of each other.

But that’s easier said than done, no? “Along the run of the strings” means you have to consider the rate of convergence of the string array towards nut in order to accurately shape the fingerboards so that it will be straight “along the run of the strings. But most factories and luthier’s don’t do that. They operate on the assumption, the fretboard is a tapered surface section of a cylinder rather than a tapered surface section of a cone. The latter though is the true case, not the former.

Factories subscribing to the former, arch their fingerboards blanks on a shaper using a cutter with a semi-circular cutting edge. This imparts a uniform cross-sectional curve along the entire length of the blank (making the fretboard surface a section of a cylinder). However, after the fingerboard is tapered and the strings are placed so that they converge towards the nut, the result is that these same strings will see a slight “hump” in the board. This effect becomes also apparent after luthier’s sand the arch into the fingerboard using a curved sanding block stroked longitudinally up and down the length of the fingerboard blank. The result is a puzzling hump, or “backbend” in the fingerboard which inevitably appears after the fingerboard is glued to the flat neck-shaft. In many cases the string tension will iron out the problem, but if the neck is too stiff, the hump will persist and dog the luthier—and the eventual owner of the guitar.

Those who ascribe to the latter visualization (a surface section of a cone)--which to my mind IS the proper configuration--cannot simply rely on shaping or sanding the board along it’s length, since the transversal radius should be constantly changing by a minute amount as the fingerboard widens. For us, the only recourse is to follow the preliminary machine shaping or block-sanding techniques above, then recontouring it by hand: hand planing it using a straightedge as a guide along the “run of the strings”. The beginner will realize that this requires some advanced hand-tool skills. The recommended fingerboard hand-planing sequence is detailed in my book.

Shaping the blanks on a CNC machine should, indeed be able to create a complex curved surface with a constantly- growing transversal radius. As for me, sadly, I have no CNC machine, so I’m doomed to rely on my hand planing skills. So if you want to CNC your fingerboard, determine the splay of your string array (the spread at the nut vs. the spread at the saddle). This determines the inboard limits of the fingerboard width and taper (i.e., its outline). Then determine the minimum arch. This can vary. It usually matches the natural curve your barring finger traces when its extended. Interestingly, strong players prefer it to be slight, weaker players like it more prominent.

The bridge pin’s contribution to acoustic coupling (cont. endlessly)
>I built my first guitar this past summer and am interested in some of the articles on your website because I want to improve the quality of the next one. One question has to do with the pegholes in the bridge for attaching the strings. I have your book and have read everything I can but have never seen this addressed. The ball ends fit thru the holes to under the soundboard. Some authorities suggest slotting the holes on the soundhole side of the pegholes for string clearance. I built a little saw from a narrow jigsaw blade with no set to do this. After doing it, it became apparent that I could saw all the way so that the string lays in the sawcut instead of the peghole, thus causing the ball ends to pull up on the soundboard. I am confused, though, because all available pegs are grooved for the string. It seems to me that pulling the strings into the sawcut so that the ball is tight against the inside of the soundboard would most efficiently transfer string vibrations. Am I missing anything?

You can purchase bridge pins with or without grooves. This is for folks with different perceptions of what's going on there. My favorite bridge pin source is www.fossilivory.com where a lot of bridge pin questions are answered.

My view is that the string's vibrations are directly and adequately transferred to the soundboard regardless of how, precisely, the strings are attached to the bridge. Where else could the vibrations go? My perception is that the acoustical signal is coupled regardless of how much of the string is in actual contact with the bridge. I view it something like soldering a braided pickup wire. The signal gets through whether there is one, two or three strands touching the lead. So, from my vantage point, the only conceivable way you could have some "loss" of vibration (i.e. damping) is to encase the end of the string in a rubber plug and insert the rubber plug in the bridge. Even doing that, the effect on the tone would be slight: remember that the string’s tension, and the portion of it that makes up the signal, is mainly coupled to the soundboard through it's pressure the saddle. Now if the saddle fits poorly, that's another issue we can argue.

My perception is that concern about the shaping of the bridge pin hole and the clearance notch amounts to purely a usage issue: If carelessly done, the pins jam so you have to forcefully pry them out to remove them, or they fall out on their own, or they are ejected across the room when you tighten the strings. To avoid this problem, the string and pin can't compete for space in the hole. But the pin has to tighten into the hole.

Oh, and the ball end doesn't come to rest on the sound- board. It comes to rest on the bridge patch.

Oak for guitars?
>In the section on side bending, you suggest that Oak is a Really good material for side bending & with this in mind I have found a beautiful piece of Quartersawn Clean English Oak, with no grain run out. This is what I intend to use for the back and the sides of the guitar. There is also enough oak left for the guitar neck and headstock, but I am not sure if Oak is a suitable material for the neck and cannot find any mention of it in your book? My main concern is that although the piece left is perfectly quartersawn, I worry that steel strings will put too muck stress on the neck. Is it a hard enough timber? Can you please advise? A friend of mine has a really nice, All Oak, hand built classical guitar that is quite old and has always played well, but obviously there is less stress on the neck, as the strings are Nylon. If you consider Oak not to be suitable for a steel strung neck, can please advise of a suitable material that would complement the oak body?

Oak, as long as it is unquestionably well-seasoned, straight grained and quarter sawn is a suitable "hardwood of opportunity" for guitarmaking. It is sufficiently strong by far for steel-string guitars. Its drawbacks? Huge pores which present finishing challenges. Glass-hard end-grain and a tendency to splinter which presents carving challenges. If your tools are razor sharp and you have the proper finishing resources, these challenges can certainly be overcome. Finally oak is reactive to humidity changes, so unlike Honduras mahogany, it is chancy to use, particularly for necks. If you must, choose the straightest, most uniform-texture, most vertical-grain sample you can. Good luck!

Leave no vapor barriers under the bridge!!
>William, I'm building my first classical guitar and am ready to glue the bridge to the soundboard. You indicate that I should seal and fill the bridge prior to gluing as I will finish the guitar after the bridge is glued. Is there a problem with the gluing surface of the bridge if I have sealed and filled with shellac/pumice/oil? I plan to French polish the guitar. I will use Titebond as the adhesive. By the way, great book! I'm having a lot of fun building the guitar.

Yes, there certainly will be. The glue won't wet the sealed wood, and the glue will certainly fail. I 'm sorry if the book was unclear on this point: you're supposed to fill and seal only the surfaces of the bridge that will get finished, and leave bare the surface that will be glued. I thought it would be self-evident. I would think that thorough sanding of the bottom of the bridge, and perhaps some wiping down with the solvent you used for the filler--and then re-sanding when dry would suffice to allow the moisture in the glue to penetrate it.

How would it sound if I...
One of my tutorial students writes: >I am building two Gibson jumbo guitars, with plans supposedly "smuggled" out of the Gibson factory (the X-bracing is very similar to a Martin). One with Wenge back and sides and the other Indian RW. Both will have Red Spruce tops and bracing (a former Gibson employee told me to go thick on the tops: between 0.130" - 0.140") . I am somewhat unfamiliar with the bracing, and I was thinking about experimenting a bit with one of the guitars bracing a bit, to wit: I am curious if you have ever butt-joined any of the finger or lower tone braces against the X-braces (similarly the soundhole braces against the Upper Face Brace and X-bracing) rather than than taper them ? How did/would this affect the sound - in your opinion ? Would this prevent the top from vibrating in its mutlitude of directions ? How would I need to adjust the cross-section of the bracing to accomodate this "tightening" of the bracing?

[More novel bracing proposals follow in the query]

Let me know if any of this makes sense. I know that a lot of the tinkering with bracing that people do (e.g., scalloping) does some, but relatively little to alter the sound. I am curious if, for example, combining tightening up the bracing while reducing the mass of the bracing wood might change the sound (hopefully in a good way). I would likely just run the experiment and find out for myself, but I thought that I would ask you first, if there is a liklihood that my ideas are really counter-productive; why waste quality wood, eh ?

Your email asks questions that simply, I can't answer. It is difficult to gauge the effect of one change on a guitar, let alone multiple changes. You can't expect to dream up a bracing scheme and then ask someone how it's going to "sound." I wish I could do that! I'd be writing you now from the French Riviera. Or, I could come back with some expert-sounding gobbledygook about "loosening up the vibrations" but I don't want to waste your time or mine.

Regarding soundboard design, all I can give you is what I've already given you: the outcome of my personal process of soundboard design refinement--The details of which are in my book, and in the soundboard instructions that you followed while taking my tutorial. If you have the desire to do something different, you've undertaken your own journey. I can't tell you where it will lead you. What would happen if you tied the minor face braces into the kerfing? Damned if I know. I don't do it. Do it and find out what the effect is. Or non-effect.

The only thing I can respond to the plethora of questions you have, is that they all seem to add up to stiffening up the top, and limiting it's range of response. I don't know what you or your Gibson advisor expect to gain by adding more mass or buttressing the top even more massively than the "norm". I'd advise against everything you've proposed, but would be unable to explain why in specific terms, except that it is pretty much diametrically opposed to my entire approach: which is to create an optimally minimum [i.e., optimally efficient] structure in the soundboard. I'm always asking, how can I brace this more efficiently? Not, what would happen if I beefed this up? Or increased the top thickness by 30%?

I'd also be extremely skeptical of any Gibson insider counseling massively thick soundboards. These may be the same voices that nearly drove Gibson to ruin after they choked the market (mainly by selling them on Gibson's past laurels) during the seventies with hundreds of thousands of awful, terrible acoustic guitars that, while massive and crude in construction, nonetheless collapsed in all kinds of unrepairable ways.

William Cumpiano's String Instrument Newsletter #22

My soundboard looks like a potato chip!

William Cumpiano's
String Instrument
Newsletter #22