Official Luthiers Forum!

While I'm no expert, it appears that Martin uses the 1/2" saddle height on most (possibly all) their guitars (meaning distance from soundboar to top of saddle). Who knows how that came about (an interesting discussion) but it seems many luthers have carried that standard forward and use that saddle height in their models. I've even heard claims that 1/2" is the "optimal" saddle height. This intrigues me.

What makes the 1/2" saddle height optimal? Because Martin does (or did) it that way while valid, isn't an argument for optimal. Seems to me that the "optimal" saddle height probably varies with the stiffness, bracing pattern, etc. of a given top.

The bridge rotation due to string pull is a substantial force on the soundboard and requires more stiffness to resist than resisting the direct pull of the strings with no rotational force. The moment (the rotational force) the strings place on the soundboard is directly proportional to the saddle height.......so half the saddle height and you half the rotational force. Since this force wants to "kink" (for lack of a better term) the soundboard, the final top thickness and bracing height are likely dictated by this force. So what if the saddle height were lowered where the rotational force was closer to the force required to resist string pull? Seems we could reduce the overall stiffness of the top where it's more responsive.

That "kink" that the rotational force wants to put in the soundboard........seems it would tend to dampen at least some vibration modes on the soundboard possibly changing the attack and/or sustain. If you lower the string height, couldn't you also decrease the bridge weight? That seems to be a good thing (see Trevor Gore's new books). It's not fair for me to put words in anyone's mouth but I believe I've read Brian Kimsey and Mario Proulx have each stated they prefer the tone of a guitar with less saddle height and less soundboard stiffness vs a normal saddle height and the stiffer soundboard it requires (please correct me if I've mis-quoted). Seems logical to me. I realize most factory guitars sound better with more saddle height.......but they are often overbuilt so that's no surprise and doesn't negate that you might improve the tone with less saddle height AND less soundboard stiffness.

So why is 1/2" the norm? If you think this is optimal, do you mind sharing why? If you've experimented with less saddle height and less stiffness, I would be interested to hear your thoughts on the changes in sound you noticed.

_________________
Formerly known as Adaboy.......

The sweet spot is that 7/16 to 1/2 inch. Now this isn't true for all guitars but Martin style it seems to hold true. You can do the math and figure the resultant of the forces applied. There are many variables involved . Bridge plate size and all but the key I have found is that once you start getting over this 1/2 mark you can start to over torque the top. When you over torque you may also loose some tonal quality and this can over dome the top.
One thing you can do with your own guitars are to add saddle height and you can see what happens. I don't trust my ears I prefer to use an oscilloscope as the wave length amplitude shows more than ears can hear. As a Martin repair center I have seen a 000 with an over set neck. This set the bridge and saddle height about 9/16. The guitar showed heavy doming and rotation of the bridge. Once reset the bridge came back to where it belonged and the dome went down. The headroom on the guitar was weak compared to what we had after a reset and bringing the bridge back to 1/2

_________________
John Hall
blues creek guitars
Authorized CF Martin Repair
Co President of ASIA
You Don't know what you don't know until you know it

I shoot for that height as well (from reading here and other places that it is 'optimal'), but there are some FINE sounding guitars out there with less than 'optimal' string height at the bridge.

There are clearly a 'number' of factors that go into the optimizing the string geometry at the bridge (for a given guitar). I would suggest that overall string height is just one of many variables, and that we should look at it in terms of a range, rather than as a yes/no, go/no-go issue.

Just my .02

_________________
Ken Mitchell
Durham, NC

I shoot for that height as well (from reading here and other places that it is 'optimal'), but there are some FINE sounding guitars out there with less than 'optimal' string height at the bridge.

There are clearly a 'number' of factors that go into the optimizing the string geometry at the bridge (for a given guitar). I would suggest that overall string height is just one of many variables, and that we should look at it in terms of a range, rather than as a yes/no, go/no-go issue.

Just my .02

_________________
Ken Mitchell
Durham, NC

"optimal height", to me, is about playing ergonomics. it means you can dig your pick in between the strings, or get your fingertips between the strings, enough to get under them a bit, while still using the pinky, or thumb, as a reference anchor. this applies to electric and acoustic guitars. it is why i hate gibson electrics, and love fenders, because fenders mimic flat top acoustics as far as string height over "sound board". for people who only strum chords, or only use picks, i guess height would not matter so much to them. 1/2" seems about "optimal" for me personally

since 1 inch is too much and zero isn't enough lets just split the difference

_________________
John Hall
blues creek guitars
Authorized CF Martin Repair
Co President of ASIA
You Don't know what you don't know until you know it

I like to be able to adjust the action from 3/8ths inch off the soundboard to 1/2 inch so the guitar is adaptable for fingerstyle or heavy strumming. Much beyond this range the guitar looses power or over torques the bridge.

While I agree with your statements Todd, in these scenarios you haven't changed the soundboard stiffness to accomodate the new "string height over the soundboard at bridge" (thanks, that's a better way to phrase this). What if you measured the bridge rotation when the string height was at 1/2".......then lowered the string height to 7/16" or 3/8" and then played with the X-intersection height and scalloping till you had the same angle bridge rotation when the strings are brought to tension. Now you would have a valid comparison (or at least the comparison I would be interested in hearing). Have you tried that? If you based your design on this height, while building you might choose to use a bit less soundboard thickness instead of removing all the stiffness from the bracing.

Does that make sense?

Also, on my first build I started out with too high action, overcompensated and had too low action, then finally got it right. I was surprised how much the neck angle affected tone. When the action was high it was harsh and loud and it warmed up and had better bass as the neck angle was corrected. I didn't change the saddle or bridge while making these changes so that wasn't a factor. So it appears a change in neck angle has it's own impact on tone.

_________________
Formerly known as Adaboy.......

I always find myself slightly bemused by these discussions about string height above soundboard.

Surely , in order to draw any meaningful conclusions, you have to take into account (at least ) three inter-relating factors: Bridge thickness, saddle height above bridge, and string breakover angle.

Discussion about string height above soundboard per se, without considering these other three parameters, seems to me to be somewhat pointless .

I'm trying to isolate the factors causing torque on the bridge which is caused by the moment arm and the tension. We aren't changing tension. The moment arm is the bridge thickness plus the saddle height above the bridge......so yes, they have their own influence. Does that mean that the affect of changing the torque on the bridge can't be isolated? Not easily, but it seems possible.

_________________
Formerly known as Adaboy.......

I've been trying to isolate some things too, and am struggling toward getting a paper on it in to a journal. Here's what I saw in tests on a classical guitar:
1) Comparing mechanically produced and pre-recorded 'standard plucks', people were far more likely to hear a change in string height off the top than in break angle.
2) Looking at a number of variables in the recorded plucks, such as sustain, overall loudness, and harmonic content, my statistics guru and I found that the most consistent change was in the level of the second partial of the string output, which went up with a taler saddle. This is related to the twice-per-cycle tension change of the string as it vibrates.
3) We could not rule out the notion that a steeper break angle could transmit a bit more sound to the top, but the effect was small and there were a number of confounding variables.
4) Alterations in the static shape and vibration behavior of the top with changes in either break angle (small, if the height off the top was unchanged), or string height off the top were among the confounding variables.

At this point I'd have to say that my most reliable finding from a lot of work was that I need to do the whole thing over, and do it better.

Interesting experiment.
My contribution to the discussion is to ask, when you say people were far more likely to hear a change, was the change better, worse or neutral, and how would you tell?
That's the conundrum I face with all these scientific discussions, is how would you assign the better or worse value to any change?
Some people will say it's better, others will say it's worse.

_________________
Old growth, shmold growth!

I had a recent experience which applies to this discussion, so I'll toss it into the mix.
Last month I finished a Weissenborn style guitar with a Lutz spruce top.
Initially, I set it up with a fairly low action to minimize top distortion.
A good friend & accomplished lap steel player played it & recommended raising the action to accommodate a Beard Capo.
The new nut & saddle had to be raised about .070" to easily slide the capo under the strings.
Right off the bat, this was far & away the best sounding Weiss I have heard. The spruce top (thanks Shane!) gave it great bass (often lacking in Weiss guitars) The loudness & the sustain was much better than any hardwood topped Weiss I have heard.
I was very surprised at how much volume & sustain were lost when I raised the action.
It still sounds much better than average, but overall loudness, sustain & bass response were affected dramatically.
I'm not sure why the sound was affected. Perhaps because of the extra mass of the bone nut & saddle. Perhaps the break angle increase. Maybe the increased torque on the bridge tending to stiffen the soundboard...
I'm hoping the clever left brainers among us can provide some insight on this.

I've seen this happen on lightly built guitars. I attribute it to too much torque. As you start to get the saddle too high, the first thing to go is the musicality and sustain of the bass strings up the fretboard.

_________________
http://www.chassonguitars.com

Alan
I also am fighting to quantify this . Here are some variables that I think are interesting. The overall structure will vary from guitar to guitar as the mechanism is not the same. :

Bridge plate ( Stiffness weight and thickness ) also the foot print and relation to the torquing mechanism.
Bracing height
Top thickness
Bridge height and footprint.
Saddle height . ( I find break angle doesn't have much to do with things )
Last String fixation.

Would you agree or disagree to these variable ? I know we discussed break angle on more than one occasion . It seems the more important part is string fixation location and relation to the top of the saddle . It seems that the whole assembly of the Bridge plate , top thickness , bracing and bridge.saddle has to be looked at with the torque and the footprint from this whole assembly .

_________________
John Hall
blues creek guitars
Authorized CF Martin Repair
Co President of ASIA
You Don't know what you don't know until you know it

The last Weiss I strung up has a 21/32" string height.
I thought I'd see what it did to the top,
which is a very slight dishing only to the x,
which I saw a lot of at the Healdsburg git fest,
on non lap steel instruments.
This git has really nice bass,
plenty of sustain, the 2 treble strings are slightly lacking in volume.
I will lower the saddle and hear what happens,
someday.
I have absolutely no idea what I'm doing though,
and have wondered about this subject.
Thanks for starting it!

May have already been touched on but, distance of pin to saddle would also make a nice variable to take into account.

The angles will factor out to a degree as the connection points are what matter. If you do the sine of the angles they will add up to the same as the point start to the end.

_________________
John Hall
blues creek guitars
Authorized CF Martin Repair
Co President of ASIA
You Don't know what you don't know until you know it

There are certain models, eras and brands of instruments that have bridge pin points that are equal distances from saddle slot and that would create a constant angle to the saddle for all six strings, but, there are many other instruments (majority) that are not and are perpendicular to the center line or following shape of the bridge, creating many different angles.

theguitarwhisperer asked:
"My contribution to the discussion is to ask, when you say people were far more likely to hear a change, was the change better, worse or neutral, and how would you tell?
That's the conundrum I face with all these scientific discussions, is how would you assign the better or worse value to any change?
Some people will say it's better, others will say it's worse."

Which is why I pointedly just asked: "Same, or different?" Some people volunteered opinions, and those were all over the lot, as you'd expect.

bluescreek wrote:
"I also am fighting to quantify this . Here are some variables that I think are interesting. The overall structure will vary from guitar to guitar as the mechanism is not the same. :

Bridge plate ( Stiffness weight and thickness ) also the foot print and relation to the torquing mechanism.
Bracing height
Top thickness
Bridge height and footprint.
Saddle height . ( I find break angle doesn't have much to do with things )
Last String fixation.

Would you agree or disagree to these variable ?"

Oh boy... We could go on all day with this, and never get close to the end.

The experiment I did was designed to look at two things: break angle, and string height off the top, on a particular guitar, with all the other stuff being 'the same' as near as I could make it. Since listeners were statistically able to hear height off the top, and we found a reasonably strong physical model to back it up, I'm at least somewhat confident in generalizing that result to other guitars. Any more than that would be pushing it, IMO.

The issue is that, when you start looking at all of the variables you mention, it gets really complicated. For example, bridge mass has effects on the tone, and we could, I think, properly include the mass of pins and saddle in that. How about the mass of the bridge plate? Does that count as 'bridge' mass, or is it something separate? How would you do the experiment to find out? I can tell you that it's extremely difficult to build 'identical' guitars, even with 'identical' wood. How would you make enough guitars with different bridge plates to get a statistically valid sample? Believe me, we've talked about this: the discussions always end up with persuading Martin or Taylor to let somebody come in and make lots of measurements on, say, the next hundred OM-28s or whatever. LOTS of measurements. Then the subject of money comes up...

What it seems to come down to is that each design, and even each guitar, has to be seen as a system, and you're trying to find out the values for all the variables that wil end up giving the 'best' result (however you define it) for that system. Flamenco guitars need to use a very low bridge: they've got lightly braced thin tops that would probably 'choke' with a tall saddle, if the top stayed on. One thing the light bracing does is to 'tune' the resonant modes of the top in a way that helps give them some 'cut', which you'd be likely to lose otherwise on such a light and flexible top. The low bridge also accommodates the 'golpe'. You can't change just one thing and expect it to work, and if you change too many htings you've made a classical guitar, or maybe a bomb....

Fleck wrote:
"There are certain models, eras and brands of instruments that have bridge pin points that are equal distances from saddle slot and that would create a constant angle to the saddle for all six strings, but, there are many other instruments (majority) that are not and are perpendicular to the center line or following shape of the bridge, creating many different angles."

Right, and my experiment suggests that may not be not an issue in terms of the acoustic sound. We all know about USTs and break angle, but that's another matter. Again, there's a lot of conjecture and myth about all of this stuff, and the trick is to figure out how to isolate something so that you can test it by itself. Actually, that's the first trick: then you have to do the experiment, and THEN you have to figure out how to do it again, and do it right. The physicist Gabriel Wienrich said that the problem with science is that you're always doing something you're not good at, and that's why it takes so long.

The standard designs have been arrived at by a lot of tinkering with a lot of guitars. The famous standard models are ones that will rather reliably make nice guitars even with the unavoidable variables of woods and interacting stiffnesses that have been discussed in this thread. Since none of us can control all the mysterious relevant points, we we go with what seems to make the most guitars sound the best.
I think that to try and answer the original question you would want to examine guitars that have a great normal sound with either a taller or shorter saddle height. You would look for something else in the construction of the guitar that was compensating. Stiffness of the top is a good hypothesis, but the vibrations of the back/ sides or overall weight might be relevant.
Most of us like a blend of good tone, long sustain and loudness. Most experiments of adjusting aspects of the standard model to increase on of these virtues brings a decrease in the others.
Saddle height is definately a majo variable.