Official Luthiers Forum!

This has probably been discussed before and I apologize if this is a repeating topic.

I've been trying to decide on bridge material for my upcoming build and I'm just confused if there's any real discernible difference in tone/response from the choice of wood for the bridge alone?

Also, for replacement bridges on old guitars I try to use what the client would like, I've made them out of BRW, Indian rosewood, and ebony so far, if there's another species that makes good reliable bridges I'm open to hear it. Any thoughts on Madagascar rosewood for a bridge?

From the discussions I've seen on this, it seems that the density of the wood is considered more important than the species, that is, how much or how little weight you're adding to the top for a given bridge design. Differences in damping properties of different woods might have an effect too? I wonder about the magnitude of any effects bridge material has on tone relative to other factors though.

I've used Madagascar rosewood for bridges and like it a lot. The Madagascar rosewood blanks I have are light weight, have a nice ring to them, and look similar to Brazilian rosewood.

I plan to use ziricote for a bridge when the right guitar comes along.

_________________
Once in a while you get shown the light in the strangest of places if you look at it right - Robert Hunter

You should read the article by Harry Fleishman in the latest issue of American Lutherie

https://books.google.com/books?id=TY_6A ... es&f=false

Above is an excerpt of a book preview. On page 349 it explains a little bit about impedance mismatching between the string, bridge and soundboard which may give some insight how different bridge woods will affect the sound.
In my experience the weight of the bridge and also"Q" of the wood species has an effect on the way the sound is transmitted to the soundboard and the tone of the instrument.

I like Madagascar rosewood and use it exclusively. How much does it add to the overall sound of the guitar??? I don't know, but it sure is purdy.

BRW for me. I just like the way the blanks ring when I drop them on the bench. I like Honduran rosewood too but BRW is prettier.

_________________
Steve Smith
"Music is what feelings sound like"

What Steve just said. Brian Burns has done quite a bit of Q testing.

That said ... there is no good answer to what sounds best, since that is 158% subjective. Since the bridge is part of the whole guitar, interacting, I believe this is not a question that gets a fair answer.

That said, any rosewood or ebony would be generally considered fine. I agree with Steve Honduran Rosewood would be great, but what a snoozer. Osage Orange rings like a church bell, but heavy (and ugly for most guitars). My general choices would be Brazilian Rosewood, then Brazilian Rosewood. If that's not available, I'd look at African Blackwood, and in the Ebony category, Macassar Ebony. That said, 30 other species will work, all with those fairies dancing on the head of that pin.

I'll see if I can find it. Thanks sir. And thanks everyone for your input, I'm off to hunt down a suitable piece of brazillian now.

BZ for me as long as I can get it. Like the way it sounds and looks plus it’s got a certain swagger.

_________________
It's not what you don't know that hurts you, it's what you do know that's wrong.

I'm in the camp that holds that mass and stiffness are the main issue. I'm not at all sure about damping.

As the book says, the main function of the bridge is to define the vibrating length of the string. It does this by staying where you put it, and providing a sufficient impedance mismatch. Most of the impedance mismatch is provided by the mass of the bridge, although the stiffness comes into play as well. It also has to be strong enough so that it doesn't break, particularly by splitting out in front of the saddle, and designed so that it will stay glued down. In that regard, the actual gluing surface area seems to be less important than the shape of the footprint. In particular, increasing the width of the bridge; the distance from the front to the back edge, is probably more useful than making the footprint larger in area. A standard classical guitar bridge can have a larger footprint than a 'belly' bridge, but won't usually hold up long under the stress of steel strings.

I've used all sorts of woods for bridges, and they all work if you design them carefully. As an example, I like to use walnut for bridges on 'domestic' wood guitars, and find them particularly handy on ones with cedar or redwood tops. Both of those woods have low peel resistance, so it's especially important to see that the bridge doesn't pull up. The lower density of walnut allows you to make a bridge with a larger footprint, and wider from front to back, to keep the stress at the back edge down to within the limits the wood will abide. You can also set the saddle back further from the front edge, reducing any tendency for it to split out, without crowding the pins and ending up with too high a break angle, which just puts more stress on the front of the slot.

It's really handy to have a target weight for your bridges, and some way to keep track. The 'correct' weight will depend on what you're making. A scallop braced top, which is more flexible in the center, will need a heavier bridge to provide the necessary impedance mismatch for the strings; I find 30 grams is a good target. On a top with 'tapered' bracing you can go lighter; say 25 grams or maybe less. A lot depends, too, on the sound you're after: a more massive bridge will give more sustain and lower output, particularly in the trebles.

Bridge length and stiffness, particularly out in the wings, are also variables you can work with. This is probably more important on classical guitars than steel strings, but can provide some further flexibility.

Overall damping in the guitar probably effects the sound, but it's hard to isolate that as a variable. Material damping is even harder to look at. It's all too easy to start with a great set of material and mess it up by bad design or poor construction: factories do it all the time. Since it's probably impossible to make 'identical' guitars that sound the same, the small differences in tone that you hear from two guitars with different bridge wood are hard to sort out. I've had cases where a couple of grams difference at the bridge have made a noticeable change in the sound: if you swap out a 25 gram rosewood bridge for a 30 gram ebony one, is the loss of treble due to the higher bridge mass or higher damping?

I've picked up a lot of different woods over the years for bridges. A fair bit of it, I bought as turning squares. I'm glad I made the investment, because the price of African BW, Yucatan RW and other's have skyrocketed. I've got Mad RW, ebony, Brz RW. I've been wanting to try some Hormigo and Ziracote. So many woods, so little time, lol.

Brent

I too have tried a lot of different bridge materials including unusual ones like walnut, maple and even oak. I did once swap out bridges on one of my guitars from ebony to rosewood just to see if I could tell a difference and I didn't notice one. Of course I didn't think back then to actually weigh the bridges either, I just made them the same. Two of the classical guitars that I built I used <gasp> ebony for the bridges instead of the traditional rosewood and one of those just happens to be the best guitar I've ever made, the other one was pretty good too. Pure luck but the point being I don't have enough experience and knowledge to know what the species of wood will do to the tone of the guitar when making a bridge and I'm not sure anyone really can. So if you want tradition then stick with ebony or rosewood. Focus on the size and weight of what ever species you are using to hone in on your target tone. That's my philosophy of bridge design in a nut shell anyway.

I'm not smart/experienced/refined enough to know whether different species yield different tonal results. I can say this: Honduran Rosewood makes a great "ping" noise when you excite it, but I recently used a few pieces for bridges, and I am not crazy about the appearance. It gets these little divots that are not easy to deal with. Maybe it was just the pieces I used. I'm going back to Indian Rosewood and Ebony for a while, just based on appearance and ease of carving/sanding.

Although I agree that mass and stiffness are the main components of impedance (analogous to resistance) I still think that the wood's "Q" has a (frequency dependent) effect on the transmission of vibrations to the soundboard, similar to what is called "reactance" in electrical systems. Other components of the guitar contribute to the impedance of the guitar and quantifying how much each contributes in a finished instrument might be difficult to do.
Toward the end of the second video Mr. Burns mentions testing bridges for "Q" and his thoughts on it:

http://brianburnsguitars.com/my-process

Given the wave length of sound in wood, and the small dimensions we're working with here, there can't be any appreciable phase difference in the force at the bottom of the saddle slot and the top of the soundboard. Under the circumstances I think of the bridge as transmitting 'force' rather than 'sound'

It's certainly possible that there are losses (hysteresis) within the bridge material, but I have to wonder how large those really are. As a reality check, I once did a quickie experiment on saddles, comparing a bone saddle to one made of polyethylene; about as different as you can get. I thought I could hear a difference, but cussed if I could measure one. I know how easy it is to fool myself, so I could hardly give much credence to what I though I heard. If I couldn't pick up a difference in a measurement there, how much of a difference is there likely to be between a rosewood bridge and an ebony one, aside from the mass?

The "experts" say low density (low weight) is very important. This makes sense to me since it is easier to get a low mass object moving that a high mass object. Gore shots for bridges in the 15 to 20 gram range and includes some carbon fiber to get that low as I recall. I think Alan's 25 gram weight would work well too and allow getting away from the CF step. Gore likes walnut and paduak and does not recommend ebony due to it's high density.

I agree with what Alan says about other aspects of the bridge, especially the footprint size being important to keeping it glued down for the long term.

Hi Alan,
My thinking is it may be possible that different woods act like "leaky" band pass filters, allowing some frequencies to be transmitted easily, while impeding others. Rosewoods preference for classical guitars may be because it allows the high frequencies to be transmitted easily, while ebony is often used on steel string guitars because it damps them more and allows the bass to be heard better.
Honduran rosewood is prefered for Marimba bars rather than Gaboon ebony. Their weight is similar, but tonally they are much different.
The experiments you did with the saddle I find interesting. You probably did _hear_ a difference. I would have expected the difference between bone and polyethylene to be measurable, especially since people will argue about the differences between Tusq and other high density plastics and bone saddle materials.
I once had a waldzither with a glass bridge, which I guess is traditional for those instruments, but which is almost three times as dense as Indian rosewood. Ivory has a density twice that of EIR, but has also been used for bridges on guitars in the past (I would think the added weight would "hurt" the sound?)
Although I agree with you that "mass" has a great effect on the workings of the bridge I do think the material it is made of makes a difference.

Clay S. wrote:
"My thinking is it may be possible that different woods act like "leaky" band pass filters, allowing some frequencies to be transmitted easily, while impeding others. "

The frequencies that get passed more easily are the resonances that cause a lot of motion in the top at the bridge. The bridge material has some bearing on this, mostly, I think, through it's effect in helping to determine the mass and stiffness of the bridge. The bridge is, after all, the heaviest 'brace' on the top, and one of the stiffest.

The operative concept here is 'impedance'. Technically, mechanical impedance is the ratio of force over velocity at a given frequency. This gets complicated, but basically, impedance is low at resonant pitches, and higher at other frequencies. That, in fact, is a way to define a 'resonance'. On something like a string, that has a very simple structure, you can define the impedance at any given frequency pretty easily if you've got the math chops. More complicated things, like guitar tops, are harder to evaluate that way; usually it's just easier to measure.

If you have two structures (such as a string and a guitar top) that are joined together, looking at the relative impedance can tell you how much energy will pass across the junction. If the impedance is the same, at a particular frequency, all of the energy in the wave of that frequency will just pass from one to the other; the wave doesn't 'see' any junction. If there's a change in impedance some of the energy will reflect at the boundary, and some will pass through. Exactly how this works in any given case depends on which way the mismatch goes (high-to-low works differently from low-to-high in some respects) and how large the mismatch is.

Suppose you had a string tied to a bridge, and the two had exactly the same impedance at frequency the string is tuned to. The wave running down the string when you pluck it would simply continue on into the top. All of the energy of the pluck would be translated into movement of the top at that frequency, and you'd expect a really loud note with no sustain. Except....

The main function of the bridge is to define the vibrating length of the string; if the length of the string is not defined then the pitch is not defined either. In the case of the 'perfect' impedance match the string would not actually vibrate at the desired pitch; when the wave hit the bridge all of the motion would continue into the top. It would just go 'thud' because now the string length includes the distance from the bridge to the edge of the top (which edge?). In order to define the string length the bridge has to provide a reasonably large impedance mismatch. It does this (mostly) through a combination of stiffness and mass. Strings are pretty light, and have fairly low impedance, even when they're tight (high stiffness). The bridge and top tend to be a lot heavier than a string, and also pretty stiff, and that provides the necessary mismatch in impedance to define the string length. At least some (and usually most) of the energy in the string gets reflected when the wave hits the bridge, and that provides much of the 'sustain' of the tone, and also defines the pitch.

As always here, we're looking for some sort of balance. A bridge that had really high impedance would give a sound with lots of sustain and a well-defined pitch, but would not allow any energy to leak out into the top to make sound. You've just invented the Les Paul. Very low impedance in the top and bridge system gives a sound with lots of attack and little sustain; welcome to the world of the banjo.

The bridge is, of course, a big actor in all of this. It's the first thing the wave in the string 'sees', so it's the point where the mismatch happens. Given the complexity of the way an acoustic guitar top moves the impedance at the bridge varies wildly with pitch. At low frequencies it's the whole system that determines this: the bridge is mostly a load on the top, and it's the stiffness and mass of the top, and it's interaction with the air in the box, that have the largest effect. As you go up in frequency things that are more local to the bridge become more important, although they may not actually over ride larger things. It's a very complex system, and that makes it hard to sort out different aspects.

Would two bridges of different materials and the same mass and stiffness sound the same? That's a hard one. It would be plausible to do a really neat experiment on that if you could build two guitars that sounded exactly the same. I'm of the opinion that this is not possible so long as you're using wood. People are not generally as good at remembering sounds over time as they think they are, and any test that can't present two choices back to back a few seconds apart is hard to credit. Nobody can swap out a bridge that fast. OTOH, it's easy to change the mass of the system quickly and when you do that you hear a lot of the sorts of changes that people attribute to different bridge (or pin) materials. I'm not saying that there is no systematic difference in sound between, say, ebony and Indian rosewood assuming the mass and stiffness could be the same: I just don't think there's any way to 'prove' it.

I think I'm going to have re-read that again a few times......

_________________
The name catgut is confusing. There are two explanations for the mix up.

Catgut is an abbreviation of the word cattle gut. Gut strings are made from sheep or goat intestines, in the past even from horse, mule or donkey intestines.

Otherwise it could be from the word kitgut or kitstring. Kit meant fiddle, not kitten.

Hi Alan,
Thank you for the post. I am somewhat familiar with the way impedance mismatching works. In my first post I linked a short excerpt from Gareth Loy's "Musimathics: the Mathematical Foundations of Music". I felt it could explain things better than my own poor verbiage.
Although it would be difficult to swap bridges on the typical flattop guitar in a few seconds, it might be possible to do on an archtop instrument. Although I realize the way they work is slightly different, still it might provide some relevant data. If I ever get one back together I will give it a try.

Arch tops and flat tops are not as different as many people seem to think they are. Still, I doubt you could swap out a bridge silently in fifteen seconds or less. OTOH, it's very easy to add or subtract mass.

If you make a jig that would slip under the strings just ahead of the bridge and raise them slightly, you might be able to slip one bridge out from under the strings and then replace it with another in less than 15 seconds. I have swapped bridges on cheap fiddles by doing something similar to that.
I don't wish to appear argumentative and I do have a healthy respect for your knowledge and opinions, perhaps the experiment is not worth pursuing, but you did say that with the saddle swap you thought you could - hear - a difference, but could not measure it. Sometimes I think we give too much credit to "apparatus" and not enough to our own sensory organs.

Clay S. wrote:
"Sometimes I think we give too much credit to "apparatus" and not enough to our own sensory organs."

It's complicated.

A big part of the issue is what I call 'Feynman's Dictum': "You are the easiest person for you to fool". Ears are wonderful organs, when they're working right. The problem is that they're attached to brains, which all too often hold opinions that color the way the information from the ears is interpreted. I know how prone I am to hear what I think I'm going to hear, or want to hear, and I'm pretty sure I'm not the only person who suffers from this. It's very well understood in the musical acoustics game that people will home in on any little bit of information that will bias their opinions, and you have to work really hard to exclude that.

Another thing that's widely appreciated is that there are things that are easy to hear and hard to measure, and things that are hard to hear and easy to measure. Ears are not very sensitive to fairly large changes in power if that's all that's changed, for example. It's part of the adaptation for the wide range of sound levels we encounter; the lowest level sound you can hear is millions of times less powerful that the loudest sound you can endure, but you need to be able to deal with both ends of the scale. The result is that it takes ten times as much power in a sound to seem twice as loud, and you have to about double the power to be able to perceive a change in loudness. It's fairly easy to measure small changes in sound power that your ears just don't pick up. On the other hand, ears can be very sensitive to the presence of sounds at low levels just above the limits of perception. It's easy to see the survival value of that. Musically, the addition of a high overtone with only 1% of the power level of the fundamental can be heard as a difference in timbre. This is one reason it's so hard to make 'identical' guitars that sound the same: very small differences in local stiffness of the top wood, say, can alter the high frequency output of the guitar enough to be perceptible.

The ear/brain system has evolved to extract an amazing amount of information from the sounds around us. A normal room will have hundreds of resonances in the high frequency range, which can cause sounds reinforce or cancel out in certain spots. Moving a microphone a few inches produce a radically different sound spectrum from a given input, but the ear/brain system deals with this fairly easily. The output of a typical telephone is not exactly 'hi-fi', but you can pick up the 'phone and know almost instantly that it's your wife on the other end, and she's not happy... This is done through a lot of parallel processing, the application of some assumptions, and probably a lot of other stuff we don't understand yet. The problem is that there's no way to turn all of that stuff off and get an 'objective' appraisal of what the sound is.

I could go on for a long time here. There are ways to do experiments that allow you to draw conclusions about sound changes resulting from various modifications, but they're not easy, and have their limits. Again, there's lots of literature on this. The bottom line is that its hard to do good science in this field. I'm not arguing that sense information is inferior to instrumentation, but simply saying that each needs to be understood and used properly for what it's good for.

I'd like to re-enforce what Alan has said regarding the difficulty of correlating acoustic measurements with acoustic perceptions.

I've been measuring the acoustic spectra of the guitars I've been building by striking the unstrung bridge with a mallet (tennis ball impaled on a stick) and recording/analyzing with Audacity. I take a number measurements as I build, including one just before initial stringing up. The process gives a lot of technical information but not as much enlightenment as I'd hoped. I can generally identify the coupled primary resonances of the box, the top, and the back, with the spectrum invariably showing a number of higher resonances tapering to inaudibility around 3 - 4 khz. When I play the guitar before break-in, it often sounds "constrained" or 'boxy" to my ears. I then "break in" the guitar with 30 to 40 hours of mechanical strumming (enabled by a salvaged-and-repurposed automotive windshield wiper mechanism driving a wooden arm holding a pick). Playing after break-in, I invariably perceive the sound to be far more mellow and resonant, characteristics which increase somewhat in the following weeks. But when I take another spectral measurement, the main resonances have, typically, not migrated significantly. The array of higher resonances, however, has become somewhat better defined though not necessarily of greater amplitude -- certainly not a quantifiable change, though an undeniable enhancement of the tone.

I've just received a couple of "torrefied" tops (from one of the sponsors of this site) and, when building with them, will look to see if the "playing in" results in either the improvement in tone or the development of the higher harmonics that I've seen in the past -- or in both.