Builders: relationship between top, back, and air?

[Luca Canteri]

In his books Ervin Somogyi is always and constantly repeating "this is what I prefer, there are different ways to do this and none of them is right... all of them can produce a very good intrument..."
I like to blend different ways in a constant search of my own.

[Burton LeGeyt]

Bruce,

What exactly is nonsense? I am not sure what you are so against here, especially since you go on to say that you too consider the relationship between the top and back important. Is it just using numbers rather than fingers to make the distinction? I think most people, even the ones who promote the science as loudly as you promote the instinct, treat the data as a means to an end (a great guitar) and not the end in itself.

[Bruce Sexauer]

The nonsense I refer to is the idea that the analytical mind can negotiate the engineering problem represented by the difference between the mundane guitar and the "great" guitar on a consistent basis through measurement based on Chladni patterns, taptones, pitched voicings, phased plates, Heimholz frequencies, bowed edges, etc etc. The problem is more complex than that, and the solution is simpler. I consider all of those input devices a distraction, they are therefore literally able to hold back true progress which I believe lies in the intuitive self despite the common belief that progress comes from the intellectual analysis. I think I am answering the OP's question. I am not against anything, merely saying how it seems to me.

[arie]

first off, thank you bruce for taking the risk in addressing this post. it is appreciated.

second, imo I think that the end result is all that matters. does the guitar sound good and play well? regardless if one consulted tarrot cards or chladni patterns, used a pocket knife or a NC mill, etc... will the public buy it? is the client happy?

for me i think a mix of both worlds (scientific and intuitive) is what i like. i don't want to really spend a long time in a lab with a sound generator taking notes nor do i always do something because "it's always been done that way" either. i think that a good grasp of the big picture (results) is what matters for me.

[justonwo]

That an engineering understanding of guitar construction holds back progress in developing guitars is, in itself, "complete nonsense." Just because a system is complex does not mean that it can't be explained in reasonable engineering terms. Guitar tone is not produced by magic or mojo. It's produced by physics . . . regardless of whether or not the luthier happens to understand those physics. In many cases, it may be far easier to rely on rules of construction garnered through experience and/or intuition, but at the heart of all those rules lies a physical explanation.

Having read Ervin's books, I believe they provide reasonable explanations for most key design variables in a guitar and an excellent framework for understanding how guitar tone is influenced. To say that experience and intuition is the only methodology for pursuing an understanding of guitar construction is, I believe, a defeatist position prone to the creation of mojo-based crackpot mythology.

Of course, I've never built an acoustic guitar so I'm sure that nullifies my opinion entirely.

[gitnoob]

I'm a big fan of science and engineering, but I see an analogy with cooking. You can have a great understanding of chemistry, but that doesn't necessarily make you a great chef.

[Matt Mustapick]

Originally Posted by Bruce Sexauer

The nonsense I refer to is the idea that the analytical mind can negotiate the engineering problem represented by the difference between the mundane guitar and the "great" guitar on a consistent basis through measurement based on Chladni patterns, taptones, pitched voicings, phased plates, Heimholz frequencies, bowed edges, etc etc. The problem is more complex than that, and the solution is simpler. I consider all of those input devices a distraction, they are therefore literally able to hold back true progress which I believe lies in the intuitive self despite the common belief that progress comes from the intellectual analysis. I think I am answering the OP's question. I am not against anything, merely saying how it seems to me.

Bruce is making a subtle point that is difficult to put succinctly, though I think he's done a good job of it. By my reckoning, the point is well worth appreciating. I can't say it better than he has so I won't try. I don't mean to suggest that Bruce and I would agree on everything when it comes to this issue, but his words here expresses my own thinking precisely.

[Bruce Sexauer]

I am in no way anti-engineering. My point is that to consistently achieve greatness, engineering (intellectual understanding) simply isn't enough. Because each piece of material is at variance the combination of parts its too complex to manipulate through quantification. Many have tried, and results max out at high mediocrity; very good perhaps, but only occasionally stellar. Literally all of the makers I most highly respect for consistent greatness (there are but few) have great latitude in their concept and construction for subtle non-technical manipulation. That "something" which cannot be quantified but must be included, an undefined variable.

I completely disagree with the statement that there is no mojo or magic in guitar-making. While there doesn't have to be, I am aware of no great builder who does not use it. Of course I am defining "great" by my own rules, but I am judging the instrument, not the maker or the process here. And when I see a large proportion of what I consider to be great instruments coming from one source, I do jump to conclusions. I do call that which I believe in but have no hard evidence for mojo/magic in this case, but what I actually think it is is getting the intellect out of the way so that the mind behind, the sub-conscious mind, can control the process. This 2nd mind is the holder and processor of all experience, and unimpaired by "my/our" input is capable of things that appear to be magic. I don't call myself mojoluthier for nothing, I mean it most earnestly, even if it is partly a goal and not yet fully actualized. It is absolutely where I'm going. While guitars are actually tools, the process of creating them can be art.


Again, this is my story, not necessarily typical, but the OP asked. In other words, IMO! Additionally, I believe I am onto something, and I may be guilty of proselytizing some. I would like to sway the odd luthier away from the production end and toward the art end of this craft as I believe it is a better life, and often not considered.

[Matt Mustapick]

There's an assumption happening here that is the source of a lot of confusion, and it's this...
- If a method employs strictly quantified formulations, lookup tables, and such, then that makes it more like science and engineering.

This would be true of the quantities and formulations were sufficient to tackle the problem. But if those quantities and formulations are very insufficient to capture the essence of the challenge, then this approach can be mostly a distraction, taking one's pursuit further from the mark than another approach that relies more on tactile and along-the-way sensory data, and reliance on comparison to past experiences, without any reliance on attempts to model total systems.

Take a moment to consider the complexity of the system of a vibrating object like a guitar...with all of its wildly variable inputs (energy from the player) and indeterminacies (uncontrolled variables that even the most exacting modeler and craftsman would have to concede). How would a vibrating plate be affected if the rim was .5 mm narrower than spec? ...if the kerfing were 2mm less tall after the rim had been dressed to meet the back? ...if the side of the kerfing that will be glued to the top is at a 2 degree angle with respect to the relaxed top, so that when they are glued together a subtle stress is imparted onto the top? ...if the back were thinned an extra 4% in one area in the process of surface prep? Can such a vastly complex and indeterminate system be modeled with math that is not absolutely blindingly complex? No...no way...fuhgetaboutit! Could it be a mistake to rely primarily on an oversimplified model as a lamppost to guide exploration? Of course.

I applaud any effort to bring any intellectual idea to bear in the pursuit of a better guitar, and that certainly includes these kinds of modeling/quantitative types of approaches. Perhaps thinking that way can help builders come up with new ideas to guide their methods. But I'm skeptical that modeling and quantitative methods can form the essential basis for understanding guitars and creating successful recipes. It's possible to analyze Tony Hawk's skateboarding technique with motion-capture and computer modeling, and in some context it *could* be worthwhile as compared to other ways of striving, but I *definitely* would not recommend this as a line of inquiry for anyone who wanted to gain an appreciation for great skateboarding. In my opinion, so it is with guitars and guitar building.

Lastly, this is just my point of view. It's not meant to be taken as a pointed rebuttal to anyone's outlook. If anyone happens to see things another way then that's probably a very good thing in the grand scheme.

[gitnoob]

As the OP, I didn't mean to suggest that treating the guitar as a simple 3-oscillator system was a sufficient model.

I'm intrigued by the idea of an "active" coupled back vs the back as a reflector, so I wanted to understand that a little better.

The most obvious example of an "in-phase" back would be a back with the same resonance frequency as the top. I thought surely that's not what Somogyi meant by "in-phase," and it turns out he doesn't mean that -- like most other builders, he builds the back to be stiffer than the top, but he doesn't quantify the relationship in his book.

He leaves interesting clues, though.

My post wasn't about the "best" approach. I'm just curious about how builders approach/view the relationship of the three primary oscillators.

[Matt Mustapick]

Originally Posted by gitnoob

My post wasn't about the "best" approach. I'm just curious about how builders approach/view the relationship of the three primary oscillators.

I guess my previous post was a long-winded way of saying that I don't approach or view that relationship at all, because I distrust the idea that it's a useful way of understanding guitars. As it happens the backs of my guitars are much much stiffer than the tops. That's not a design aspect of my guitars. It's how guitars are made. It's just a happenstance of the most mundane tonal and structural necessities of the craft.

Rather than think about these relationships, I think about how the last guitar I made sounded, how I feel about it, what I might like it to be different, and then I take an educated guess about how I can tweak the design. I do that over and over every time I build a guitar. So far, that process has not involved any assessment of the top/back/air relationship because that relationship, and the way it functions in a guitar, is so intractably, unfathomably complex that I haven't viewed it as a practical consideration.

My process has involved assessments of many other things... How flexible should the top/bracing system be? How does it affect tone if the bracing/top system is kept the same but the top is left thicker/stiffer while making the braces smaller/more-flexible? Vice-versa? What areas of the top should be thinner than other areas? I only mention a few of these considerations so that I won't end up leaving the impression that I don't think there's anything worth examining. I examine things that seem comprehensible. If something is so vastly complicated that a serious analysis of it would require Fourier transforms, multi-variable and vector calculus, nonlinear partial differential equations, stochastic methods and the like...well I'd have to start with an algebra and pre-calc review, so that'd be a *really* long row to hoe if I was to take it seriously.

I guess I've gone off on a bit of a tear here...it's not meant to be an airtight point of view. It's just sort of a very general indication of "where I'm coming from" with regard to the OP's question.

[justonwo]

Originally Posted by Matt Mustapick

There's an assumption happening here that is the source of a lot of confusion, and it's this...
- If a method employs strictly quantified formulations, lookup tables, and such, then that makes it more like science and engineering.

If you look at Somogyi's books, which I think is what we're talking about here, this is not the assumption. I don't think anyone would attempt to deconstruct guitar building into a set of lookup tables or equations. Somogyi's books do not present equations and lookup tables. They provide some physical explanations/theories that can be used as guideposts for understanding guitar making.

One interesting example from a Somogyi lecture I attended. In addition to talking about the vibration modes of the top, he talked about the cube rule of load bearing beams. Put simply, the stiffness of a load bearing beam (i.e. brace) goes as the cube of its depth. From a practical standpoint this means that small changes in brace depth can have a huge effect on their stiffness.

Somogyi did not then suggest that this rule should be plugged into some kind of "guitar top modeler" to yield brace location or shaping. It's just a simple engineering principle than can help guide thinking.

The same was true of the vibrational modes. With some basic explanations of vibrational modes, he was then able to talk about how different bracing structures can produce different sounds/qualities by enhancing or inhibiting different vibrational modes. Again, no complex math or lookup tables were used. Just some relatively simple physical explanations.

If Somogyi can produce consistently spectacular results (pure speculation on my part, based on the comments of others and his revered place in the luthier world) using a quasi-engineering/physics framework - and furthermore if he can transmit that understanding to other builders using the same framework (through his classes and books) - then it is clear that it can be done. That doesn't mean it has to be done. Obviously, there are some builders here who produce great results without physical abstractions to guide them - I can speak to the quality of Matt's work firsthand - but that doesn't mean those explanations/theories/etc are nonsense.

[Matt Mustapick]

If the incorporation of basic engineering principles is the extent of the quasi-engineering approach then there's probably less disagreement in practice than in description. Bruce knows and I know as well as Ervin or anyone else that a thick stick is a whole lot stiffer than a thin one. It's possible though that the introduction of exponential notation into the discussion might lead some to imagine that this line of inquiry could lead to greater precision in determining guitar tone...third degree quadra-nomials predicting overtone amplitudes or what-have-you. I think that a quasi-engineering way of talking about guitar building can be suggestive that there is some sort of sophisticated approach going on behind the scenes. That would not be a useful understanding of the state of the craft, or how it's successfully accomplished. The simple fact is that guitars are not engineered, not like that anyway. A recipe gets tried, judged, and tweaked, but never truly understood. Repeat as needed. Engineering principles are brought to bear, but not wielded so as to analytically synthesize an intellectual understanding of the entire system. There are builders barking up that tree, but in my view the tree is *extremely* tall, and some of the barking is emboldened by an under-estimation of its height. That's the only point I'm trying to make. And I think Bruce was heading in a similar direction.

[Burton LeGeyt]

Thanks for the clarification Bruce. And thanks also to Matt for enlarging the discussion (and for using a sweet Tony Hawk example).

I can appreciate where you guys are coming from but having spent the time to learn how to examine (with numbers) a few aspects of the guitars I build I can't imagine not checking them and seeing what they might tell me. It is no different than what Matt is doing, I change things slightly on each build. I don't feel I am doing myself a disservice by using numbers instead of straight muscle (and regular) memory to recognize the changes but I am willing to accept that I might be.

It is interesting finding myself not in the complete instinct side of this issue. My brother is a physicist and I went to art school, we would have these kinds of discussions all the time with me never on the science side! I guess he swayed me a little. And he is the kind of person who could rock some skateboard tricks and may well have done some calculations off of watching someone do them to figure it out. Oh well.

I approach guitar making as tool making and try not to think of it as art, I am somewhat protective of that word but I don't discount the, for lack of a better word, essence that each builder imparts on their instruments. I think that is something innate and it is not something I ever thought to just trust and go with. That would feel to me like a big (and somewhat narcissistic) leap of faith. Maybe though it does need to be practiced to fully mature and give that extra something to the process. I can't imagine it would need your complete attention but again, who knows? Maybe that complete attention is what really pushes it over the edge.

[Matt Mustapick]

Originally Posted by Burton LeGeyt

...thanks also to Matt for enlarging the discussion (and for using a sweet Tony Hawk example).

I can appreciate where you guys are coming from but having spent the time to learn how to examine (with numbers) a few aspects of the guitars I build I can't imagine not checking them and seeing what they might tell me. It is no different than what Matt is doing, I change things slightly on each build. I don't feel I am doing myself a disservice by using numbers instead of straight muscle (and regular) memory to recognize the changes but I am willing to accept that I might be.

It is interesting finding myself not in the complete instinct side of this issue. My brother is a physicist and I went to art school, we would have these kinds of discussions all the time with me never on the science side! I guess he swayed me a little. And he is the kind of person who could rock some skateboard tricks and may well have done some calculations off of watching someone do them to figure it out. Oh well...

That's a gracious post...cheers.