Counter point to Somogyi's saddle design philosophy?

[mcbean]

Hey guys,

I'm not a builder and am pretty ignorant, just to be up front about it. I've been reading various articles on Somogyi's website and one of the things that I'm really curious about is his philosophy of saddle design as articulated here: http://www.esomogyi.com/principles.html

My question: If this wider saddle inherently increases the efficiency of the string contact, why has this design not become more prominent? In other words: Is his philosophy an incontrovertible truth? Is it merely one way of increasing efficiency (of which there are other ways to do it while maintaining a traditional "low efficiency" saddle)? I know, of course, there are many factors that go into the efficiency of the guitar...but is there a school of thought that suggests the smaller saddle, coupled with other features, can somehow be more efficient than what Somogyi proposes?

If this is simply the "best" way to design a saddle- why is this not ubiquitous in the industry? Is it just taking a while to catch on more widely? I know a number of high-end builders (even those that aren't Somogyi students) use this saddle design, but curious as to why it hasn't funneled down elsewhere.

Thanks in advance,
Scott

[Bruce Sexauer]

I do not use a wide saddle because it is not necessary and it adds weight to the most sensitive area of a guitar.

[mcbean]

Originally Posted by Bruce Sexauer

I do not use a wide saddle because it is not necessary and it adds weight to the most sensitive area of a guitar.

Had not thought about weight! That makes sense. Thanks for chiming in.

[redir]

I've used both methods and tend to stick more with the standard 1/8th - ish size now but you really can intonation a wide saddle really well, that's it's advantage. Somogyi is up against hundreds of years of tradition and like anything in guitar making this is just, as The Dude says, "his opinion man." So it's hard to make anything a gold standard in guitar making. He has his reasons just as Bruce stated his reasons, that's just the way it works and will continue to work.

[MC5C]

It's my belief that as long as you don't exceed the elastic limits of the material and cause significant deformation from the vibration, the area of contact has very little effect on how effectively energy is transferred. Point contact, wide contact, all will transfer the same as long as the material (bone, wood, metal) is not deforming. Hence archtop bridges with tiny little adjuster screws bedded into wooden bridges transfer vibrations actually quite effectively. This is similar to how friction works, surface area of contact isn't in the equation for friction. Area doesn't count until the energy level causes deformation or degradation of the material (tearing, over-heating, etc).

[mcbean]

Originally Posted by MC5C

It's my belief that as long as you don't exceed the elastic limits of the material and cause significant deformation from the vibration, the area of contact has very little effect on how effectively energy is transferred. Point contact, wide contact, all will transfer the same as long as the material (bone, wood, metal) is not deforming. Hence archtop bridges with tiny little adjuster screws bedded into wooden bridges transfer vibrations actually quite effectively. This is similar to how friction works, surface area of contact isn't in the equation for friction. Area doesn't count until the energy level causes deformation or degradation of the material (tearing, over-heating, etc).

Interesting! Thanks for sharing. Very helpful.

[charles Tauber]

Prior to the advent of "modern" science, philosophers sat around and conjectured how the natural world worked. For example, philosophers concluded that heavy objects fall to the ground faster than lighter objects. That seems logical, and that's what people believed for centuries. Science of that time was based on logical argument - belief - rather than on objective observation of physical phenomenon.

"Modern" science began with Galileo creating physical experiments and observing, and analyzing, their results. For example, by dropping objects of varying sizes and weights from the tower of Pisa, he observed that the rate at which objects fall is independent of their weight, thereby disproving the centuries-held belief to the contrary.

One way of looking at the situation is that humans hold various, largely unsupported beliefs about the causes (and effects) of observable phenomena. If, or when, the conclusions of valid experiments "proves" the previously-held beliefs to be invalid, the new conclusions replace the old beliefs. Until such time as that happens, the old beliefs prevail.

There remain many aspects of natural, physical phenomenon to which the principles of modern science have not been conclusively applied. That is, for those things, rigorous, controlled, repeatable physical experiments have not been performed, the results analyzed and conclusions derived that explain that particular behaviour. In areas where that has not been done, that leaves individuals to believe whatever "seems reasonable" as the cause of observable behaviours. That is based upon "belief" or "opinion", something different than repeatable, verifiable "proof".

There is an enormous amount about instrument making, in general, and guitar making, specifically, that is unknown. These areas of the unknown remain areas of conjecture and belief. In these areas of the unknown, different individuals have different beliefs, different theories about how things work and what causes what effect. Is one person's unproven theory or belief better than another person's? Sometimes probably, sometimes probably not.

That brings us to Mr. Somogyi's article. There are some parts of his article that are based on proven, valid science - that is, based on repeatable, valid experimentation with valid conclusions. There are other parts that are based upon belief, without "hard" science to back them up. Sometimes, beliefs are valid even though there is no hard science to verify them. The difficulty is that without that rigorous process of validation, it is difficult to determine which are valid and which are not.

Part of the "clue" in his article is his use of language. He is often very clear where he is presenting valid science and where he is presenting his own opinion/belief. For example, the statement, "I think you’ll find this second guitar works better." The words "I think" belay an opinion, a belief. Similarly, the statement, "I’ll be very surprised if you don’t find that the better ones are just more responsive when you do something as simple as tapping." I'll be very surprised is not the language of valid science: it is expressing an opinion.

The difficult part in reading any technical article is in separating objective "fact" from subjective "belief". In some cases, the author might, himself, not be clear of what is fact and what is his own belief.

I've used 3/16" and 1/4" saddles since 1977. I use them strictly for the purpose of allowing sufficient room for intonation purposes. I'm not aware of any scientific work done on string/saddle contact that supports his statements regarding "high-efficiency string contact". As far as I'm aware, there is some "common sense" to not having string contact that is knife-edge, but beyond that, it appears to be his belief based upon his experience. There is nothing wrong with presenting one's belief, and a reasoned argument for holding that belief, but that doesn't automatically make it "true". Without valid support, it is one person's belief, possibly no more valid than anyone else's belief. Wars have been started over different beliefs.

"An art is a science with too many variables." P.N. Adams

"It is a capital mistake to theorize before one has data." Arthur Conan Doyle

[Howard Klepper]

From the last time this was discussed:

http://www.acousticguitarforum.com/f...6&postcount=20

[Alan Carruth]

It is interesting that Trevor Gore also seems to be of the opinion that there's a benefit to having more string contact as it wraps around the saddle, but he has not offered any objective demonstration of this either, so far as I know.

Weight at the bridge, and particularly in the saddle, is a two-edged sword. On the one hand, in the 'ideal' case the bridge and saddle 'should' have very high mass and stiffness, so that the vibrating length of the string is well defined. The problem there is that an acoustic guitar with a perfectly immobile bridge would be inaudible. As with so much else having to do with the guitar, the trick is to find the proper balance that will enable the string to work well and still produce the sound we want. That's likely to be different for every guitar.

It's really hard to know just what property of the saddle is most important. It's mass, stiffness and loss factor could all have a bearing. Is a piece of fret wire set into the bridge top as good as a full bone saddle, assuming the weight of the system is the same? It's hard to say.

Most of us seem to agree that a bone saddle is better than a plastic one. I just weighed a few saddles; the plastic one weighed about 1.5 grams, and a bone one of about the same size came in at 3. Does the bone work better because of the extra weight, or in spite of it? How about the fully compensated bone saddle for a 12-string that is 1/4" wide and weighs almost seven grams? In that case there's plenty of energy in the strings to overcome that mass, and the full compensation sure helped the sound on the high frets.

Is it going into a walnut bridge that weighs 18 grams, a rosewood one at 25 grams, or an ebony one that tips the scale at 35?

Let's not even get into pins...

As Charles Tauber says, at this point it's pretty nearly all opinion. It would be fun to do the experiments some time, for those of us who have such peculiar notions of 'fun' and the boatload of time it would take.

[tadol]

Unfortunately, I think one has to start this by defining what is meant by "best" -

Best for tone? adjustment? transmission? appearance? I'm sure another half dozen options could be put on that list by a real luthier - and that still is only dealing with one part of a larger system, that needs to be evaluated as a whole. You can't separate the bridge (shape, size, thickness, material, mass, placement), the top(thickness, material, mass, stiffness), the bracing(size, placement, shape, material, glue(?), stiffness), the pins (weight, material, or even if there are any) from the larger picture.

Most luthiers have very specific thoughts on how all these items, and many others, affect the instrument - and they each have their own focus on what is "best". They also have to maintain a focus on what will sell - which has been the source of many other discussions, and may be quite separate from what they may feel is best in terms of the instrument. Thats a contradiction that science has not really been able to rationalize -

Science is a wonderful thing, but there are some areas where it simply cannot go in any meaningful way - this may be one of them -

[runamuck]

I love your post, Charles Tauber.

I also want to add that not everyone wants a "responsive guitar": there are plenty of people who prefer heavily flatpicking a Gibson.

[Alan Carruth]

tadol wrote:
"Science is a wonderful thing, but there are some areas where it simply cannot go in any meaningful way - this may be one of them -"

Properly used science can be a great help in this area. The problems are that it's very difficult to do good science on this stuff, while there has been some bad science (or bad stuff that's made to look like science) done that ends up confusing poeple.

It has been said that 'no oscilloscope can tell me what good sound is' and that's true as far as it goes. I can think of examples where poeple have tried to 'scientificallly' determine what a 'good' instrument should do from first principles, and come up with things that are, if not actually bad, at least not usably better then what's out there now.

On the other hand, there are a few studies that have tried to correlate listener and player inputs as to what's good or not with objectively measurable properties of the instrument. None of them has defined what a 'perfect' instrument should do, nor have they tried. What they have done in offer some insight as to how the structural properties of the instrument relate to specific aspects of tone, which can help a luthier make better instruments.

These useful sorts of study all have one thing in common; they take a lot of time and effort to do. Of course, the poeple who have the money to finance these sorts of study don't need the information they provide, and the folks who do can't afford them. Thus most of the best work on guitar acooustics has been done in university engineering labs by grad students who are learning how to do vibration analysis for later application to things like cars and airplanes that pay better. We benefit incidentally.

There are a few hopeless crazies out there who actually waste time trying to do good science on their own. It's a lot easier now with the computer revolution than it used to be. Still, getting any sort of clear and useful results depends a great deal on picking out a question that is simple enough to be solvable by simple means, and it's not always easy to know that going in. Guitars are engineering nightmares, and human response to sound is worse. One thing it offers is job security, since nobody will find all the answers in a lifetime. Now it it could only be made to pay...

[Truckjohn]

I would suggest that its worth considering that Mr. Somogyi sells guitars for upwards of $30,000 each and there are no shortage of takers..... Before you throw too many rocks - consider that he is one of very few successful luthiers who has been around for as long as he has...

And so instead of saying "Oh... You can't do such and such" when its clearly being done successfully - perhaps its useful to consider how its working as part of his guitars. Because they are very good.

[charles Tauber]

Originally Posted by Truckjohn

I would suggest that its worth considering that Mr. Somogyi sells guitars for upwards of $30,000 each and there are no shortage of takers..... Before you throw too many rocks - consider that he is one of very few successful luthiers who has been around for as long as he has...

And so instead of saying "Oh... You can't do such and such" when its clearly being done successfully - perhaps its useful to consider how its working as part of his guitars. Because they are very good.

As I read it, so far in this discussion, all responses have been of a technical, objective nature. No one has "thrown rocks" (I.e. made personal comments aimed) at Mr. Symogyi.

Likely those in the discussion are well aware of what his guitars sell for, how long he has been doing it and his popularity. Those don't alter the nature of science and its requisite burden of proof.

Do you have anything of an objective nature to add to the discussion?

[Trevor Gore]

Originally Posted by Alan Carruth

It is interesting that Trevor Gore also seems to be of the opinion that there's a benefit to having more string contact as it wraps around the saddle, but he has not offered any objective demonstration of this either, so far as I know.

No, he most certainly is not of that opinion!

Some background: Alan has done a lot of experimental work investigating the effect of the break angle of the string over the saddle and has found that (cutting a long story short), break angle has little effect on the sound or power output of a guitar (please correct me Alan, if I’m misrepresenting your work here). Similarly, when I do a simple (rigid body mechanics) mathematical analysis of the break angle issue, I come to the same conclusion: break angle doesn’t enter into the equation, so to speak. Except that I hear a difference, have always heard a difference, have changed things on numerous guitars and always hear the same kind of difference, and it’s not related to saddle height, which precipitates a different set of differences.

So what’s going on? Clearly something that doesn’t show up on a simple analysis. The hypothesis (note: hypothesis!) that I put forward is that it is to do with the bending stiffness of the string and the end conditions of the termination. If a string goes over a proverbial knife edge saddle before disappearing into the bridge pin hole, the string can rock like a see-saw on that knife edge as it vibrates with a corresponding motion of that bit of string between the saddle and the bridge pin hole. The end condition of the string can be termed “simply supported” or pinned. On the other hand, if the string passes over a wide saddle with a large break angle, a corresponding long line of contact, (a substantial wrap angle) the string can’t rock where it passes over the saddle, it has to bend. And (according to my hypothesis) it is this different end condition that is responsible for the difference in sound.

A fixed end condition (approximated by a saddle with a large wrap angle) results in a different overtone structure and greater inharmonicity of the partials due to the extra damping and stiffer termination of the string. That can be demonstrated by comparing the maths of string vibrations with different end conditions (see the book) where a pinned end condition is more representative of the way my guitars behave (yes, I’ve done the analysis and testing). And this correlates to the difference I hear; pinned ends (no “wrap”) have less inharmonicity and a greater number of audible harmonics. Which type of termination is preferable is up to the listener. I prefer the sound of the “knife edge” saddle, though I use wider saddles than most so I can get the intonation I want. Wide saddles and “knife edge” are not mutually exclusive. Of course, how much “knife edge” you can create and use depends on the saddle and string materials. Soft saddle materials won’t support a knife edge, whilst hard materials that will will also damage the strings. So there is a compromise to be made there. But I usually aim for the sharpest “sensible” edge.

There’s quite a lot going on in the contact mechanics of the string termination that is not well understood and has not attracted much research. For me, I know what to do to reliably create the different sounds and I’m not really bothered about the details of the physics, because I doubt it would change what I do.

Happy to hear any other hypotheses, provided they come with at least some supporting rationale!