Compensated Saddles

[Gordon Currie]

Originally Posted by charles Tauber

The bottom line seems to be that most players either don't care, don't hear the inaccurate intonation or just accept that poor intonation is the norm - it doesn't have to be.

All of my guitars but one have had extensive work done to get the intonation as good as possible. One in particular had acceptable intonation, but I was able to get it DEAD ON. I can play intervals at the 15th fret that sound as in tune as those on the 3rd fret.

Once I experienced that level of intonation, it was impossible to ever settle for less.

[Mirosh]

Anyone ever make a saddle with extra width/depth/? fore-and-aft at the top - extra room for adjusting intonation by filing and sanding? It would still have the right thickness to fit into the slot. A typical saddle doesn't have much to work with.

[Brucebubs]

Originally Posted by Mirosh

Anyone ever make a saddle with extra width/depth/? fore-and-aft at the top - extra room for adjusting intonation by filing and sanding? It would still have the right thickness to fit into the slot. A typical saddle doesn't have much to work with.

There is another alternative - the compensated nut.

[BrunoBlack]

Originally Posted by Brucebubs

There is another alternative - the compensated nut.

I’ve wondered about this. Once you fret a string, doesn’t it take any influence of compensation at the nut out of play? Seems like you need saddle compensation to influence fretted strings. But I’m no expert.

[TRU]

Charles, you are a wealth of information, and very generous with your time.
It seems to me that if guitar techs were like doctors, then one would have a sit down with them to discuss/"measure"
1) how strongly one presses the strings to fret
2) what capo is used
3) how tightly is the capo pressing the string

Am I wrong?

I am a singer/songwriter...not Tommy Emmanuel.
I play some cowboy chords, sure. But I also de-tune to open tunings during my sets. I capo EVERYWHERE, every with open tunings, and even when playing some simple slide. And forever it has amazed me that even when I have my tech set-up my guitar, when tuned perfectly in standard, it's sharp when I place a capo at fret 3 or 5.
My entire career this has driven me CRAZY!
Your thoughts, please.

[hubcapsc]

Mike Weems made this saddle for my cf-100...



-Mike

[rcolosi]

First, Charles' post was excellent. Everyone should take the time to read it.

Here are some more facts:
- From the early days when acoustic guitar production began in increasing numbers, up until as recent as the late 80's most factory-produced guitars did not come with compensated saddles. Each guitar was slightly different. Different enough where one "set" configuration of compensation was not applicable to instrument 1 through instrument 1,000, 5,000, 50,000...whatever. Angling the saddle on a well-built "as mathematically close as possible" guitar.... just like you all have on your guitars at this very moment was sufficient enough to get it in the ballpark and make everyone happy. Happy enough for Jimmy Page's intro to Stairway...played with a guitar most people would laugh at nowadays and an uncompensated saddle. Not sure I hear to many intonation issues on that piece up at the 7th fret. I could go on and on with names, but I'm trying for brevity here.

- As the guitar production grew larger and larger, manufacturers were tailoring their production skills such that one guitar was much closer in spec to the next one and the next one and so on. Computer-aided devices for design and construction greatly enhanced this ability.

- With each instrument being so similar in spec, a standard compensation could be applied to further "sweeten" the intonation (for lack of a better term). It was now worth the time for these companies to pony up the money to pay a company who mass-manufactured injection molded plastic to design new tooling to make a billion of these things. ...So they did. It's actually more expensive than you may think to do that.

- Now, most guitars have them. Vintage recreations often do not...because they are faithfully copying something that was built in, say, 1941, in the case of a Martin Authentic. Same with many Gibsons. If those or the real deal vintage guitars are built well and/or still in spec from a scale/mathematical standpoint...the uncompensated saddles work just fine.

So yes, it was a very clever idea that was part of the natural evolution of just about any product. Is it important? You and your particular set of ears and the wooden box you have in your lap can be the judge.

Originally Posted by TRU

So, if compensated saddles are so clever and important, do folks with 100% vintage Gibsons and Martins have tuning/intonation issues that folks with trendy tusq saddles do not?

(fyi, i know very well what a compensated saddle is and its purpose)

[mawmow]

When I got my brand new little Gibson L-1 1928 Blues Tribute, the luthier had to work the truss rod and told me these Gibson had an intonation issue saying I had to lenghten the scale at bridge : I just filed the bridge nut enough to get the string/nut contact as far back as I could and it worked ! In fact, the luthier had not lowered the action enough for me, but filing the nut at bridge a bit did it too.

I do not see how tusq vs bone would influence scale lenght.

By the way, my readings over the years lead me to understand that frets intervals is a compromise as not every interval is perfect for each note : the compromise is to make almost every scale sound right. Some scales may not, and some chords could sound weird though... So, the path to perfect intonation has its own weird ways.

Add on : Thanks to Brucebubs , saddle, yeah. Sorry.

[Brucebubs]

Originally Posted by mawmow

When I got my brand new little Gibson L-1 1928 Blues Tribute, the luthier had to work the truss rod and told me these Gibson had an intonation issue saying I had to lenghten the scale at bridge : I just filed the bridge nut enough to get the string/nut contact as far back as I could and it worked ! In fact, the luthier had not lowered the action enough for me, but filing the nut at bridge a bit did it too.
.

The bridge nut is called the saddle.

[charles Tauber]

Originally Posted by mawmow

I just filed the bridge nut enough to get the string/nut contact as far back as I could and it worked ! In fact, the luthier had not lowered the action enough for me, but filing the nut at bridge a bit did it too.

By having the breaking point as far back as possible on the saddle, you've lengthened each string as much as is physically possible for the location and thickness of that saddle. Assuming the saddle was angled, you have provided a linear approximation of compensation - that is, that the amount of compensation that each string needs lies along a straight line. Usually, that isn't very accurate, but can improve intonation, depending upon the situation.

Martin is well-known to have made guitars in the 1970's where the bridge/saddle was in the wrong - too short - position, giving the guitars poor intonation.

Also, one of the factors that contributes to how much compensation an individual guitar requires is the height of the strings above the fingerboard (i.e. action). By lowering the strings, you reduce the amount of compensation required.

I do not see how tusq vs bone would influence scale lenght.

Just to be clear on terminology, scale length is usually defined as the theoretical vibrating string length, the length used to calculate the position of the frets. (This terminology isn't universally applied, adding confusion.) Once the frets are placed, the scale length is not changed by where the nut and saddle are placed. Where the nut and saddle are placed changes the actual vibrating string lengths. Hence, the choice of saddle material has no bearing on the scale length.

Where the Tusq/bone issue enters into it is that, unlike bone, Tusq is a man-made material that can be formed into whatever shape one wants. Bone, being a natural material, must be "manually" shaped - sawn, filed, milled, sanded, etc. - to achieve a desired shape. As Bob stated, Tusq can be accurately, and repeatedly, be formed to include whatever thickeness, top radius and compensation contour one wants. That eliminates the manual work required to form bone, one at a time, into a specific desired contour.

That development allowed major manufacturers to inexpensively impliment whatever contour of (nut and) saddle they wanted. Many chose to implement a saddle shape that includes the specific string-to-string compensation they wanted.

Be clear, however, that the compensation they provide is for a very specific combination of factors that influence how much compensation is required. Those factors include string height, string materials, string manufacturing method (e.g. wrapped vs. unwrapped strings/different core/wrapping arrangements), string gauge, string tension (i.e. pitch) and so on. If you change radically from that specific combination of factors - use heavier or lighter strings, raise or lower the action... - it can affect how much compensation is required to achieve the best intonation.

By the way, my readings over the years lead me to understand that frets intervals is a compromise as not every interval is perfect for each note : the compromise is to make almost every scale sound right.

That is equal temperament. On most guitars, frets are placed in accordance with equal temperament, a compromise to allow playing equally in and equally out of tune in all keys.

Some scales may not, and some chords could sound weird though... So, the path to perfect intonation has its own weird ways.

Again, intonation is one thing, temperament is another. Temperament defines the target pitches, intonation defines how closely an individual instrument achieves those pitches. Two different things.

[AndrewG]

Originally Posted by MikeB1

This is all new information for me, and I find it very interesting.

So is compensation something you never see on nylon string guitars or do you see it sometimes?

Personally I haven't come across it, but I imagine it has been done.

[AndrewG]

Originally Posted by mawmow

When I got my brand new little Gibson L-1 1928 Blues Tribute, the luthier had to work the truss rod and told me these Gibson had an intonation issue saying I had to lenghten the scale at bridge : I just filed the bridge nut enough to get the string/nut contact as far back as I could and it worked ! In fact, the luthier had not lowered the action enough for me, but filing the nut at bridge a bit did it too.

I do not see how tusq vs bone would influence scale lenght.

By the way, my readings over the years lead me to understand that frets intervals is a compromise as not every interval is perfect for each note : the compromise is to make almost every scale sound right. Some scales may not, and some chords could sound weird though... So, the path to perfect intonation has its own weird ways.

Yes, you're correct. The Pythagorean formula for calculating scale and fret positions is fundamentally flawed-the distance between nut and first fret is particularly problematic. Here's a pretty complex explanation...
http://www.stephensedgwick.co.uk/pythagoras.htm

[John Arnold]

Guitars use a tempered scale, not a just (Pythagorean) scale.
If you don't distinguish tempering from intonation, all you do is confuse the issue. Intonation of a guitar is imperfect, while the tempered scale equalizes the intervals so that all keys are 'out of tune" by the same amount.
There is no such thing as just intonation on a multi-string fretted instrument, no matter how you move the frets around. The only way that could possibly work is if you only used one chord form, and played all lead notes on one string.

[Robin, Wales]

Originally Posted by John Arnold

Guitars use a tempered scale, not a just (Pythagorean) scale.
If you don't distinguish tempering from intonation, all you do is confuse the issue. Intonation of a guitar is imperfect, while the tempered scale equalizes the intervals so that all keys are 'out of tune" by the same amount.
There is no such thing as just intonation on a multi-string fretted instrument, no matter how you move the frets around. The only way that could possibly work is if you only used one chord form, and played all lead notes on one string.

That's perfectly true John. I build and play instruments in just intonation but only fret under one string or two strings tuned to the same note. And I would also add that with just intonation fretting you are also limiting yourself to one mode on a diatonic instrument although you can play multiple modes when using equal temperament.

Originally Posted by charles Tauber

Again, intonation is one thing, temperament is another. Temperament defines the target pitches, intonation defines how closely an individual instrument achieves those pitches. Two different things.

A lot of the old mountain dulcimers I have played (and the new fretted zithers I have built myself) are in just intonation (which, just to confuse the issue, is a temperament not an intonation system!). Having practically experimented myself, I would say that it is easy and straightforward to set frets by ear pretty much perfectly in just intonation (Ptolemy’s intense diatonic scale) because the intervals are perfect – and so they are easy to find when comparing two strings tuned to the same note or a perfect 5th apart. You just slide the fret wire around until you reach the points of least dissonance (I make and fit staple frets from 18 gauge steel music wire).

IÂ’m sure that this two string fret setting system was the way many early Appalachian dulcimer makers found their fret positions as it requires no measuring, and no external reference pitch. However, you end up with frets set for a specific scale on a single string (played against root and/or 5th drones) and intonation is dependent on the string gauge, tension, nut height and saddle height the maker used when setting the frets. i.e if you change the set-up the instrument will no longer play in tune.

Making my own instruments and setting the frets by ear on some, and writing a fret calculator for just intonation fretting for others, has given me quite an insight into intonation and temperament on instruments. On old mountain dulcimers (of which I have quite a collection from many prominent historical makers) and on my own design Bocs Cân Idris (a Welsh fretted zither) I can tune the instruments by ear because the string intervals and fret placements are all “perfect”, so it is very easy to hear when a string is “in tune” or “out of tune”.

However, although I can hear “out of tune” on a guitar I can’t always hear “in tune” because, in equal temperament, everything is slightly “out”. So I find it quicker to use a good quality clip on tuner and a quick strum through open G, C, F and D to check the chord blends are “good”.

Going back to detailed saddle intonation on guitar (and nut intonation); basically, for my playing style I don’t need that level of intonation control. Particularly because I capo up the neck and so re-tune when I place my capo. I therefore never play a note on a high fret that has been tuned down at the nut. I do have a compensated saddle, but really don’t need it, and it can actually throw the b string out a little when I capo – that and the low E are the two strings that I most often have to “tweak” when I fit a capo.

A guitar with the frets set in equal temperament, with the strings tuned in equal temperament, and with the VSL, bridge and nut set for perfect intonation in equal temperament with sound OK wherever you play it on the neck – but it will never be “perfect”.

[charles Tauber]

Originally Posted by Robin, Wales

A lot of the old mountain dulcimers I have played (and the new fretted zithers I have built myself) are in just intonation (which, just to confuse the issue, is a temperament not an intonation system!).

At the risk of splitting hairs, "just" isn't an "intonation" or a "temperament" it is a system of tuning. It is "just tuning", sometimes called Pythagorean tuning.

As you are aware, just tuning relies on whole-number ratios to define pitches. "Temperaments" are schemes of small adjustments to those ratios intended to get around the inherent limitations of using only whole number ratios (i.e. the Pythagorean comma).