Official Luthiers Forum!

Very curious about techniques, priority, and methods to achieve good/excellent/perfect intonation?

I've only built a few guitars and have gotten pretty lucky with intonation and cannot really hear it when it's off anyway but have always placed the bridge as square to the string line as possible with an angle to the saddle that I read on the internet and I repeat have gotten pretty lucky or least acceptable to my untrained ear and tuner gear.

My guitars sit in my living room and I'm not getting $0000.00's of dollars for them but I would imagine if you are getting top dollar that anything less than perfect intonation is just not going to cut it.

I bring this up because I have a new designed bridge that is asymmetrical with a saddle slot cut that is cut relatively square but I can now (due to the asymmetry) angle the bridge with strings run through a dummy string holder at the heel of the guitar and over the saddle to adjust intonation at least for the high and low E strings.

As far as priority, does anyone angle their symmetrical bridges/saddles to improve intonation? Not saying a crazy angle but something that no one would ever see or notice to improve intonation?

Is there a science that gets it done easily and accurately? Split saddles? Perfect saddle angles? Wide saddles for adjustment?

_________________
Thank You and Best To All

All the science you'll need is in the books by Gore and Gillet. Expensive, but worth it. I can give you a brief paraphrase here.

Intonation has to be understood in light of the scheme of temperament you choose. All temperaments are compromises, and have deficiencies, which once they become apparent, never go away. Guitars use 12-Tone Equal Temperament by default: it's the only scheme that uses straight frets because it's the only one in which all of the semitones are musically the same size. Many intervals in ET are 'out of tune', in the sense that they will have significant beating. The big musical advantage is that, since all the keys are 'out of tune' in the same way you can modulate freely from one key to another without the need to re-tune or re-fret: they'll all sound equally good, or bad.

When you press a string down on a fret the string is stretched, and the pitch rises. Fret locations can be calculated and the slots cut to any desired degree of accuracy, but string stretch will mess up the intonation if you set a guitar up that way. In theory you can compensate for this by relocating the nut and saddle and correct the intonation nearly perfectly. In practice there are a few more variables that are hard to account for.

To a first order of approximation the amount of compensation needed at each end of the string will depend on the properties of the string itself, as well as the scale length used to calculate the fret locations and the action height. This assumes that the string is always pushed down with the same force, which, of course, it is not, and that there is no relief in the fretboard.

A bigger problem is that the calculations used assume that the ends of the strings are 'fixed'; in particular, that the bridge doesn't vibrate in response to the string forces. This is, of course, not true: if the bridge and top didn't move the guitar would produce no sound. All acoustic guitar tops do move, more at some pitches than others, and in so doing they can alter the pitches of particular notes or partials of notes. In general, the better and more responsive the instrument the more of this sort of pitch change you'll see.

The bottom line, then, is that it's possible to get arbitrarily close to 'perfect' intonation for most notes on most guitars by properly compensating the nut and saddle. Given that some intervals are already pretty far 'out', you don't want to make that any worse by having poor intonation. It's well worth the effort to get it reasonably close.

Thank you so much, Alan, for the reply. Curious, would you sacrifice a perfectly square bridge to achieve better intonation, at least at the 12th fret where most of the interest seems to lie?

_________________
Thank You and Best To All

Intonation is a subject that requires a fair amount of study and thought. The reason for that is there are many ways to approach it. Everything from just sloping the saddle and calling it good to installing intonated nuts (see pic), intonating the saddle, cutting a little off of the first fret, installing fragmented frets and saddle pieces and few others that escape me at the moment.

The goal of intonation is to get the equal tempered fret board to play in tune as you move up and down the neck. It is no easy choir mainly because the strings reside above the fret board and must be pushed down to the fret to chance the note. A higher set up will not be intonated the same as a lower one. A player that tends to push the strings to the side slightly while fretting a note will take a different intonation setting, etc.

The way I approach that is what I consider the simplest. This is an abbreviated description.

First I remove just less than the width of a fret from the end of the fretboard, then I locate the saddle so that the distance from the 1st and 6th string positions at the saddle and face of the nut pretty much match StewMac's fret setup page. (I have developed my own calculator that is close to StewMac's)

Next comes stringing the guitar up, and doing a setup (truss rod adjustment, action setup etc.) I play it in awhile and tune it every couple days for about a week.

Then I measure the intonation comparing an open note with a note fretted at the 12th fret. I use a Peterson Strobe tuner. If it's off I adjust by changing the break point at the saddle for that string until it plays in tune.

_________________
Joe Beaver
Maker of Sawdust

Standard steel string bridge already sacrifices symmetry for the sake of intonation (angled saddle slot, versus the straight one on classical bridges). Fan frets need enough saddle angle that the bridge itself usually needs to be rotated or twisted in some way, but for regular frets you'd have way too much saddle angle by the time the bridge needs rotated.

But actually with nut compensation as described in Gore/Gilet, you need less saddle angle than usual. Think of it as taking a standard setup, and then moving both ends of the strings closer to the 12th fret, with low E and B moving the most, and G and high E moving the least.

Larry,

If you need to put your bridge off square, perhaps your saddle slot is not angled enough. You also have the width of the saddle slot to play with by adjusting the break point as you shape the saddle

Thank you all for the replies. I have been lucky so far to have the saddle slot angle just about right with the bridge set up square (as I can get it) to the string line but was wondering if it ever became necessary would any of you angle the bridge just a little one way or the other (saddle slot angle remains constant) to improve intonation? I guess I'm really only talking about the 12th fret for the sake of this conversation and having something that changed the previous intonation setup.

And on this latest bridge design, it is not symmetrical, meaning I can move and change its angle (saddle slot again remaining constant) to get the best 12th fret intonation with a string holding jig on the heel and the look should work no matter what the end angle of the bridge ends up. Pretty sure this will work but my failure rate with ideas can be pretty high.

Again curious if anyone would shift the angle (slightly, slightly, slightly) of the bridge to improve intonation?

_________________
Thank You and Best To All

There was a thread where a user suggested that changing the bridge angle slightly to improve intonation was a small sacrifice. Can't find it now.

The more I think about this the more unreasonable the idea seems as in order to really duplicate the actual set-up it seems there needs to be an actual set-up not a mock up that might even be different when finally strung up.

Doesn't mean I won't experiment, but may mean the experiment might fail...

_________________
Thank You and Best To All

Oh, you mean if you already have the bridge made with saddle slot cut at some approximate angle, and then it turns out you need it rotated a bit to get the intonation just right? Then yes, of course intonation takes precedence over visuals. But you shouldn't get yourself into such a position in the first place. Either figure the exact right angle for the slot when making the bridge, or route it after the bridge is glued on.

I used to pre-rout my saddle slots, but now slot the bridge after gluing it on.
I just position the bridge square using SM fret calculator measurements, allowing for saddle compensation.
Then I route the saddle slot after the bridge is glued on by positioning a maple strip (as a movable temporary saddle) under the E strings and adjusting it's placement to get good intonation at the 1st and 6th strings.
Once I'm happy with this false saddle's position, a drop of thin CA at each end secures it, then use a saddle slot routing jig to rout though the maple strip to depth.
Fit the saddle, and shape it for good intonation on all strings.

_________________
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.

Thanks Dennis, hard to read everything in to these posts but you answered my question and I agree with your last sentiment...

_________________
Thank You and Best To All

Nice - seen it done that way before and a bit "router on the bridge no turning back" shy and have seen it done with dremels that I hate but could set it up with real router no problem I guess. Will keep experimenting..and thanks so much for the reply.

_________________
Thank You and Best To All

I'm not keen on the Dremel either, nor the plastic SM jig, so built my own.
I was a bit "router on the bridge no turning back" shy myself for the first couple of times, and still double check (several times)
But with the weight and power this is solid and stable so I'm getting more confidence in it.
Even had a student use it very successfully, although I did set it up.
He made the cuts and depth adjustments as he routed.

_________________
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.

I never had a problem after I mounted a proxxon motor on the Stew-Mac router base...

Nice jig Colin..Thanks for sharing that.

_________________
Thank You and Best To All

Dennis K wrote:
"But actually with nut compensation as described in Gore/Gilet, you need less saddle angle than usual. Think of it as taking a standard setup, and then moving both ends of the strings closer to the 12th fret, with low E and B moving the most, and G and high E moving the least."

Actually, the nut is shifted closer to the 12th fret, and the saddle is shifted away. Since the pitch goes a little sharp (usually) at the first fret you need to move the fret back to lengthen the vibrating part. Since you'd end up having to move all the fret back by about the same amount, it's easier to just shift the nut forward and drop the tension on the string a smidgen.The nut does end up close to parallel to the saddle, and Gore talks about simply splitting the difference. That is, if you normally cut the saddle slot with a 1/8" slope from treble to bass, you can cut both the nut and saddle with 1/16" of slope, parallel to each other, and with half the offset at each end, and get reasonably close to 'correct'. Again, almost anything you do within reason will tend to be better than not compensating at all.

I rout the saddle slot first, and then position the bridge carefully. I also use a wider slot, which allows for correcting the peak to get the intonation more accurate. The 'correct' break line on the saddle top is almost never straight. In particular, it's common to need more compensation on the B string than the G, and the angle from the high E to the B is different from the angle from the G to the low E. Some folks use a split saddle for that. With 12-strings, which require different compensation for the octaves, you need at least a 3/16" wide saddle.

But when you move the nut toward the first fret and drop the tension a bit, that means all fretted notes will now play flat compared to the standard setup, since they're affected by the tension drop but not by the nut movement. So the saddle needs to come closer to the 12th fret to sharpen them back up.

DennisK:
and then I wrote:
"The nut does end up close to parallel to the saddle, and Gore talks about simply splitting the difference. That is, if you normally cut the saddle slot with a 1/8" slope from treble to bass, you can cut both the nut and saddle with 1/16" of slope, parallel to each other, and with half the offset at each end, and get reasonably close to 'correct'."

That is, both the nut and the saddle are cut off perpendicular to the center line of the strings/neck. The nut is closer to the first fret than it 'should' be in theory, with the low E end being moved in more than the high e. The saddle is moved back, away from the 12th fret, but less than you may be used to if you have not been compensating the saddle, and with less slope. I thought my explanation was clear, but apparently not. Mea culpa.

Hmm, maybe we had different initial conditions in mind? My explanation was starting from a standard setup, where you have ~6mm saddle compensation on the low E. If you were starting with the saddle at the exact theoretical scale length, then yes, it does move farther from the 12th fret than that.

If you have ever watched a dog chasing it's tail then you would know that is a lesson in 'perfect' intonation.

I understand. As far as I'm concerned the best way to discuss changes is to start from the theoretical calculated positions, and correct from there. You might assume a particular setup that is different from what somebody else would do, and end up really chasing your tail...

And then there is the nonluthier approach:

www.youtube.com/watch?v=V2xnXArjPts

Interesting discussion. When I'm setting intonation I setup the guitar to play by putting a 1/8" rod across the bridge before any slotting is done for the saddle. I can turn and tilt the rod until I get good intonation according to my tuner. The advantage with the rod is that the guitar is actually playable with it being held down by the string tension. And I can mark the saddle lines fairly accurately because the rod gives some clearance on the bottom.

Any small adjustments I can make by cutting some intonation in the saddle.

VERY interesting...

_________________
Thank You and Best To All

Nice technique...

_________________
Thank You and Best To All