Tuesday, February 22, 2011

Are luthiers fetish-makers?

|ˈfeti sh |
nounan inanimate object worshiped for its supposed magical powers or because it is considered to be inhabited by a spirit.a course of action to which one has an excessive and irrational commitment : he had a fetish for writing more opinions each year than any other justice.a form of sexual desire in which gratification is linked to an abnormal degree to a particular object, item of clothing, part of the body, etc. :Victorian men developed fetishes focusing on feet, shoes, and boots.ORIGIN early 17th cent. (originally denoting an object used by the peoples of West Africa as an amulet or charm): from French fétiche, from Portuguese feitiço ‘charm, sorcery’ (originally an adjective meaning ‘made by art’ ), from Latin factitius (see factitious ).

Lies, damn lies and acoustics I

What makes a guitar sound good?

For the interested musician or dedicated guitar-making student who seeks to get to the truth about the instrument's nature, the scientific and the popular media have, alas, provided only a muddy minefield.

The field of guitar acoustics has become populated both by sincerely interested speculators who, with little fear and with remarkable facility, pass on untested information as if it were gospel; and a few real experts who usually confine their comments to the scientific literature.

In the scientific journals and professional publications, you will find that when researchers make even the most modest observations, they make them with guarded hesitation--elaborately qualifying them with an eye to the strict rules of the scientific method. They will make their observations about the guitar's behavior only from what is clearly evident from laboriously-generated data. Not a sentence more, not a period less. Thus, guitar enthusiasts searching for enlightenment in the turgid prose of formal acoustical papers (even those that can actually follow them), usually come up with a dry hole.

Guitar makers, on the other hand, have no such hesitation--all too often, I catch them making shameless claims about their guitars in the popular media--purposely allowing the impression to remain in the public mind that the keys to the mysteries of guitar acoustics are well in their reach. Perhaps that, too, is to be expected: many of them must find it irresistible to create impressive sounding, though ultimately vacuous, impressions of their products. After all, who will know the difference?

It is difficult to sort out the wheat from the chaff in this matter because even instrument makers themselves (and who else can we turn to?) have created and perpetuated a body of acoustical mythology that pervades the background of acoustical inquiry, like noise. Instrument makers, for all their intuitive skill, usually make very, very poor sources of hard information on guitar acoustics. The vast majority is not trained in acoustics and I'd hazard to say that only a handful could actually define "sound" or explain the behavior and effect of sound waves traveling in air or through solid objects. How many are driven by a profound--or even working understanding of point sources, sound pressure, resonance theory, acoustic declension, near-field holographic interferometry, acoustical damping or standing waves? Hardly a one, myself included.

In truth, they should only be asked about what they do best: create charming cultural artifacts by following patterns and practical strategies passed on to them by their teachers, duplicating the work of past masters whose work--marvelously accomplished, skilled and intuitive work--offers up a treasury of effective problem-solving solutions. To this they add the wisdom gained from their own painful trial and error past experiences--and then, each adds subtle variation to this acquired methodology, the product of their personal inclinations and individuality.

As if this wasn't impressive on its own merits, I nevertheless see too many "guitar authorities" ducking difficult questions by wrapping themselves in the mantle of vague acoustic-science terminology, gulling the un-initiated among us into thinking that they have the sound thing all wrapped up.

For instance, a prominent guitar-making textbook put forth a popular misconception among traditional builders here and abroad: that, starting at the bridge, the guitar creates sound by waves "radiating" away in all directions toward the edges of the guitar, not unlike the effect that follows the dropping of a pebble into still water. That ignores, of course, the fact that the string is coupled to both ends of the guitar, the much-ignored neck as well--making it a significant sound source at certain frequencies. In reality the guitar top--nay, all the guitar's surfaces--wobble and heave in all sort of deceptively fitful ways, in response to the energy interactions occurring between the guitar and its vibrating strings. The explanation also ignores the fact that the guitar's vibrating wall surfaces account for only a fraction of the its total sound production. Yes, the guitar is a fantastically complex vibrating object which defies simple description--and simple descriptions like "sound emanates from the bridge like a pebble in water" can only be misleading at best, cynically self-serving at worst. 

Yet, I have read in guitar magazines solemn assertions by experienced guitar makers and observers that sound waves travel to all areas of the soundboard via its braces. I call this the "highway" model. For all intents and purposes, though, braces are invisible to the several major low-frequency guitar vibration modes that have been identified on a guitar (there are too many high-frequency vibration modes to individually identify). Then there is the "bass-treble side" model which divides the soundboard into discrete bass and treble halves (each "under the control" of the corresponding bass and treble strings immediately above them). In this scheme, alterations to the bass and treble response of the guitar can be directly manipulated by somehow fussing with or changing the thickness or height--or number--of soundboard struts in the corresponding area. They talk like they can put treble and bass controls on an acoustic guitar by fiddling with the "bass and treble sides" of the soundboard. My acoustician friend tell me that's a ridiculous myth, too.

We have also seen the soundboard likened to a trampoline, or requiring its individual parts to be tuned to the notes of a chord so that the instrument's response will take on the sound character of that chord, be it sad or happy or mellow or bright. I kid you not.

These acoustical models are passed on in what is claimed to be a good faith effort to educate the public. But I suspect, many times its done to maker the bearer of such "wisdom" seem wizard-like. Either way, these silly claims fly in the face of even the most elementary teachings of vibrational physics. But that doesn't stop the copy writers in the guitar media. For example, in various trade magazines one advertisement boldly claims, "...scalloped, hand tapered 'parabolic' braces not only give an even balance between bass and treble, but give the clarity of tone ordinary bracing cannot achieve." Yikes! Another ad for a mass-market import boasted that the edge bindings on their guitars are made from "natural Maple--this greatly improves tone and acoustic qualities. The rosewood bridge patch transmits more vibes."

So how should a luthier respond to the question of how guitars work...truthfully?
How about,

"the guitar is essentially a cultural artifact which has evolved in form over the centuries in correspondence with the aesthetic preferences and technological level of its times. Essentially, it is a complex resonating system that transforms the kinetic energy of its strings into acoustic energy in the air surrounding it. Very little is known for sure about how it does that. It offers a method for the user to control just the fundamental frequencies of the harmonic components of the acoustic energy of the generated sound field--that is, the pitch of its notes. The builder determines the production of the remaining frequencies--the ones that determine the quality and character of its notes--only to a very limited extent. However, builders can do so only after persistent trial and error and after developing a refined intuitive sense. But the lion's share of the remaining frequencies in the sound field are fixed by cultural factors such as the guitar's accepted form and by the sum of the intrinsic elasticities of each of its many components."

Although essentially accurate, this description still gives an unsatisfying answer. So a wise skeptic ought rather to view the guitar-maker's work more like a painting or sculpture than as an electronic sound device without batteries. That would be closer to the mark.

Thursday, February 17, 2011

Quarter sawn wood vs. rift-sawn vs. vertical grain wood

The adjective "quarter sawn" is popularly used to describe boards that display its annular rings at 90 degrees to its widest face. Much goodness flows from this vertical grain orientation, particularly in ways that are important to luthiers: stability, stiffness, structural efficiency and on and on. However, in truth, there is actually very little "quarter sawn" wood which actually has vertical grain! So if you go to a lumberyard looking for "quarter-sawn" wood, expecting to find vertical grain boards--expect to be disappointed.

(image credits: http://www.wisegeek.com/what-is-quartersawn-wood.htm)

The "quarter sawing" strategy is so-called because the log is cut into quarters and then the quarters are all sawed at a 45-degree angle yielding...very little actually vertical grain lumber!!
That's because the purpose of quarter-sawing lumber is not to extract the greatest yield of vertical grain boards. Not at all. The purpose is to maximize the yield of planks with straight-line figure on their wide faces.

Plainsawing yields the greatest yield of lumber, period. It also yields the widest boards. But it produces very few straight-grain-figure boards and very few vertical grain boards. It produces the greatest amount of wavy-grain-figure or "cathedral" figured boards.

Quartersawing yields more waste that plainsawing. It yields more vertical grain boards than plainsawing, but narrower boards. But all the boards it produces are straight-grain-figure boards.

Now look at the rift-sawing diagram. It would appear to yield 100% vertical-grain boards, but with the greatest amount of waste of all. Indeed some sawyers have computerized equipment that rotate the log to maximize vertical-grain boards, leaving much of it however, on the floor as waste. Technically and correctly, that is called rift sawing.

But this is weird: there is widespread confusion as to what rift-sawing actually means. Commonly, "riftsawn" boards are generally considered to be boards...with the annular rigs oriented 30 to 60-degrees--essentially, most anything between flat and vertical. So that is the term of art, although the popular term is technically incorrect.

So much confusion. What to do? Easy. Stop using the wrong and confusing term"quartersawn". Don't specify "quartersawn" when you ask for luthier timbers: specify "vertical grain".

PS. You all know how particular I am about soundboards. I don't go to lumberyards looking for vertical grain spruce, no matter how "vertical" it seems. "Vertical grain" spruce boards are an inferior source of soundboards for luthiers. If you are a follower of the school from whence I came, you only get your soundboards from suppliers who process spruce specifically for stringed instruments--that is, they don't saw planks at all. They buck the logs at predetermined lengths, use wedges to split them apart into blocks and then saw the soundboards off the split faces. Calling the product of this process "quarter-sawn" soundboards is non-sensical. Calling them, say, "split" soundboards is rather more accurate.

Wednesday, February 16, 2011

Where the rubber hits the road

I am in the process of finishing a OO 12 fret steel string. When I adjusted the neck angle by adjusting the heel/guitar body joint I ended up with the correct angle per your book page 304 but a gap between the end of the fingerboard and the sound board of 0.060 - 0.080". I suspect this is not uncommon but I am not sure of the best way to deal with this. Just clamp the finger board to the soundboard when gluing the neck in place or shimming the fingerboard extension with an ebony wedge?

This is precisely where the rubber hits the road, vis a vis your building chops. It should all come together: your bridge height and scale length; your fingerboard thickness, how upright your headblock was when you glued your back on, and if your upper transversal face brace had sufficient proper arch. If all those are working precisely and in concert, your fingerboard end will just come down and sit on your soundboard nicely when your neck angle is cut appropriately to your bridge.

The gap you describe is really the total amount of slop that accumulated in all of the above.

If you put a wedge under the fingerboard end, it would look screwy. And that would contravene the Prime Directive: It Can't Look Screwy. On the other hand, you could just glue the fingerboard end down, the most expedient and simple solution. No one will notice, no harm, no foul. It won't look screwy.

But if you try to play beyond your 14th fret, you would probably find it impossible. On a regular guitar, you could say, well, it'll be okay as long as no one plays beyond the 14th fret. But believe me, someone is going to try to play past the 14th fret sooner or later, and then the luthier Gods will punish you.

So what's the solution with the immediate problem at hand? Well if you're lucky, there's the following fix: reset the neck angle for a thinner bridge. That will bring your fingerboard-end down. Were you planning on a 3/8" thick bridge? Well, you're in luck. 3/8" bridges are dogs. 5/16" thick bridges usually work and sound better for anything below Dreadnaught, and even on a Dreadnaught they're better most of the time. Resetting your neck back up is going to move your nut closer to the body, but you can just make up for that when you set your bridge for glueing.

Did you already set the neck angle for a 5/16" bridge. Oh well, that invisible fix is not available. Sorry. You may have to resign yourself with the expedient recourse of just glueing your fingerboard end down to the soundhole.

Next time: make sure your headblock is not allowed to float when you glue the back on. It's real easy to have it accidentally rotate a tiny bit when you clamp the back over it. So make sure it is immobilized 100% when you glue the back on.

If it was immobilized, then the chances are that for your scale, you had insufficient arch in your upper transversal face brace. Make note for the next time to add extra arch.

Also make sure your workboard is flat and rigid if it is cantilevered off the edge of the table. That will throw a chunk of randomness into the equation.

Ed.: The fellow reset his neck for a 5/16" bridge and solved his problem!

Sunday, February 13, 2011

More thoughts on wood-specie choice

I'm repeatedly reminded by my experience and research that the choice of wood species plays a rather small effect on the the quality of tone of the guitar made with it.

Excuse me while I duck.

This is an extremely controversial statement in some quarters! I've had to avoid some rhetorical stones thrown at me from makers persuaded to discriminate between a far more narrow list of traditional choices than I.

Regardless, my own perception is that you'd have to posses the ear equal to a spectrographic apparatus to unmistakably discern between a rosewood and a mahogany guitar, let alone between a brazilian rosewood and a thai rosewood guitar. The myth that certain woods are "nobler" than others has been gladly perpetuated by the trade, and as you would guess, the "better-sounding" woods tend to gravitate towards the top of the cost and rarity scale. But I've found over and again, that cost and rarity of the tone woods does not a great guitar insure.

Some recent evidence: I've recently discovered Michigan Sycamore. A student sent me a few sets from a tree that fell in his back yard. Several guitars I subsequently made with that wood rank up with the best-sounding guitars I've ever made. So much for cost and rarity! And the color and figure is as exotic and visually inspiring as the finest Brazilian!

Some older evidence: flamenco guitars. Spanish guitarmakers built Brazilian rosewood guitars for their professional trade, mahogany and walnut guitars for their student trade--and cypress guitars for their holes-in-the-shoe-soles trade. Time was, you could throw a stick in any direction in Spain and hit a cypress tree. The Spanish Roma were the dispossessed population and cypress was literally the cheapest building material you could obtain. The two were joined in flamenco guitars and the rest is history.

Cypress is a conifer--a softwood--a light, foamy-textured gymnosperm, like spruce. Rosewood, of course, is an angiosperm--a dense, glassy hardwood. Totally different fiber and vessel orientation. Nothing more contrasting in density, tambour, texture--you name it--than cypress. Rosewood had to be brought on galleons from Brazil, swapped for barrels of Port wine from Portugal, then sold in Spain as imported lumber or as furniture that was later deconstructed by luthiers to make high-end guitars with. Cypress was sawn from trees in the guitar maker's backyard or cousin's field, where for centuries, long lines of cypress trees divided fields up and served as windbreaks.

Clearly the choice of woods for the fabled flamenco guitars was not made to please guitar "connoisseurs"--it was made to satisfy the pocketbooks of the low-income populace. Yet for at least two centuries, the Spanish Roma created legends of musical history with these knotty, figure-less, "expedient" woods.

I look upon flamenco guitars as guitars with soundboxes made entirely of soundboard material. Is it thus an instrument lesser in grandeur of tone than their rosewood cousins? Is their tone dimmer? Is their musical status inferior? If you believe that, you must listen to a flamenco guitar some day.

So this guitarmakers advice would be to cast away historic prejudices in this regard. Widen your horizons as well as your experience. Go forth and save the rain forests. Check your own back yard. It's safe.

Are larger guitars necessarily louder?

I have not found any real-life evidence to support the popular assertion, "larger guitars are louder guitars." Larger guitars seem simply to have an expanded bass range, usually at the expense of its treble response. My acoustician-mentor Tim White maintained that smaller guitars express their acoustic power in the voice range, to which the ears are more sensitively tuned. That very well may explain why many well-made small guitars can seem to be extraordinarily loud. I can demonstrate that effect with several sub-0 size (also known as parlor or "Ditson") guitars I've made in my shop with students. These 13-14-inch guitars are becoming very popular nowadays, ever since, seemingly, Sting appeared on YouTube with his "baby Ditson". They are ever bit as loud and satisfying to play as any good, large guitar--perhaps even a shade louder. I find the trick is to keep the string scale the same length as you would on a larger guitar. The old parlor guitars tended to sound weak and wimpy, I feel, because their small size was always paired with sub-sized string scales. Pair up a parlor-sized soundbox with a Dreadnaught scale, and you've got yourself a potential small cannon--I've found, and repeatedly.

My soundbox goes dead right after routing for bindings

The book you wrote with Mr. Natelson is an invaluable source of informations and suggestions, as well as your website and blog. Let me say that, besides the strictly technical guitarmaking teaching, I do appreciate your radical intellectual integrity, refusing any kind of shortcut in the making method and in particular in the "skills building" process. Yes I know, this way everyting is a bit harder, and sometimes discouraging... well, for these and other reasons: thank you. ok, ok... let's get in topic...

After routing the binding ledges (7 x 2 mm) I noticed the tone of the soundbox had dramatically changed. this seems reasonable to me, since the routing process removes material at the nodal point of the oscillation, allowing a larger vibration (is it right?). after binding and purfling installation the intial tone was almost recovered, however with substantial changes. so the question is: do you think that the binding choice (material, dimensions) and installation can affect in a relevant manner the final tone of the instrument? If not, can you imagine I did something wrong in my binding installation process?

I honestly don't know for certain. My guess is that since routing the binding usually reveals a number of gaps or holes at the corners that permit air into and out of the soundbox, usually at the spots where the transversal top and back braces meet with the sides (and especially when individual tentellones are used), this creates a lot of air leaks around the rims. The principal of an ordinary resonator requires a single point of escape for the air which is trapped inside the soundbox, and if you pierce a resonator of any sort with countless small openings, all the acoustic pressure gets dissipated instead of focused. Thus--I theorize--the soundbox goes dead when you tap it after routing the perimeters off. A resonator shot full of holes is no longer a proper resonator.

Apparently, binding the guitar subsequently plugs all the small gaps and openings around the rim, and the guitar now can clearly resonate all the frequencies that are are excited when you tap it. I don't wish to speculate as to what the ultimate effect is on the finished guitar's tone. However, it's fair to say that the soundboard has changed in a fundamental way from that of a clamped plate to that of a hinged plate after it has been bound and it is thus freed to vibrate at a considerably lower frequency and greater amplitude than it could before the binding mortises were routed. I imagine that's why the soundbox sounds louder, clearer and deeper-pitched when you tap it after it is bound. No doubt scraping the bindings reduces the thickness somewhat also around the rim, freeing even further the lower-frequency response of the entire soundbox.

Disclaimer: I have no proof of any of this nor, I'm sure, has anyone else. I've only described my reflections on the subject. I can already hear the sounds of skeptics claiming that I'm all wet because guitars with multiple soundholes seem to resonate just fine. Okay, fair enough: I would hasten to ask these folks what happens when you route the edges of these multiple-soundhole guitars, does the same deadening drop in resonance occur? Their response would be revelatory: it would enlighten us as to the differences between the resonance of a given soundbox with one full-size soundhole, one with several smaller soundholes, one with a second sizable soundhole cut out of the ribs (a feature gaining popularity nowadays)---and a series of many smaller air leaks around and just beneath the perimeter of the top and back plates--as what happens after you rout for bindings. And the results of this inquiry...alas, will inevitably lead to even more questions....

My simplified steel-string bracing design

I noticed a picture of an elegantly braced guitar top on a forum that was credited to you (heres a link to the forum if you're interested).

It certainly shows some Cumpiano characteristics, though has only 1 lower face brace and 1 finger brace each side, is it one of your later tops? Obviously it differs from the standard in this respect, and it got me wondering (if in fact it is one of your tops) how you have developed your craft over the years? For example, does a later Cumpiano guitar show any character of sound that differs from an early one? Has there been any particular character found wanting in an instrument (eg unbalanced treble/bass response, lack of sustain etc) that you have made steps in subsequent instruments to correct? From what I have read on your blogs you seem to aim for 'less weight' regarding guitar tops - has this led to any changes in bracing design striving for better efficiency and 'improved' performance or have you just refined your approach through every step required and arrived at where you are?

My bracing scheme has indeed evolved from 1984, when I wrote the book (I had been making guitars for 11 years already when I wrote the book) to now (27 years later). When I wrote the book I was closer in time to the modalities of the workshop where I learned my trade, and the way I learned to build guitars in the Gurian Workshops when I left there in 1974 remained evident in the book I wrote in 1984. Those guitars indeed were considerably heavier-braced and heavier-built in general than the guitars I've been making more recently--because over the years I have become more confident in the soundness of my own thinking about guitar design, and I have since broken away from those earlier designs in many respects. That photo you sent is indeed typical of how I brace steel-string soundboards today.

The guitars I was making in the mid eighties with the design philosophy of my teachers were more stoutly built in every regard. Their sound was pleasing and they were commercially successful. But they had a rigidity and opacity that began to trouble me, as my experience broadened and I realized that the guitars that inspired me with their surrounding, airy, crystalline, transparent sound--were without fail all exceedingly light in weight, and their bracings spare and minimal. As I studied the work of many of the past grand-masters of the craft, particularly those in the classical guitar world, I became aware of a factor that has since colored my approach, and is the philosophy that I now teach and espouse: "minimally adequate structure." It is actually a design philosophy well known to architects and bridge builders, who know very well that the most successful structure is the least: every once of material beyond that which is necessary to support the load is working to bring it down. I heard Erwin Somogy say it best: "The greatest guitars are built right on the cusp of collapse." I believe that the guitar is a loaded structure like an airplane or a suspension bridge. In the guitar's case, where minute and instantaneous changes in string tension are actually the string's signal--it's become my perception that any degree of mass or stiffness beyond that which is necessary to successfully support the string load--is working to mask or impede the signal. So I began to remove the elements of the guitar that were just there for the ride, and reducing the critical load-bearing components to their minimal dimension--as derived by trial and error.

The result is that I've achieved the goal of open airiness in the sound that I set out to find. But I wouldn't say that the earlier schemes were inferior--just that my guitars sound different now, and closer to my own sensibilities than to those of my early mentors. Let's just say that my stews taste different than those stews by the cooks that I originally learned from.

Acoustical effects of the classical closing bars

I certainly take your point about the sound of a guitar being a function of the integrated WHOLE. But in arranging fan bracing on the classical guitar, I wonder if you could explain how one decides to use closing bars or not. What part of the acoustical whole do the closing bars affect? Your book seems to take the issue for granted (installing them), while other makers have gone both ways. Can you explain what general differences come about?

All bracing designs are the result of the personal notions of their makers--or the notions of the teachers of their makers, without any science or data whatsoever to back them. There is simply no way to document--or for that matter, to express--what the effect is on the "acoustical whole" of any particular element of the guitar's anatomy. Sorry.

You may have to inquire elsewhere. But I suspect you are likely to get a less candid answer.