The fretboard must be laid out accurately - most are. Are you skeptical of the accuracy of your fretboard?
Steven Delft at https://www.mimf.com/nutcomp/ gives you a method of checking the accuracy of your entire fretboard.
You can determine the scale length of your guitar by measuring the length from fret 1 to fret 13 and multiplying that by 1.0595 x 2. You can then print a chart of your calculated scale positions at https://www.stewmac.com/FretCalculator/ (Specific charts for popular guitar scale lengths are also provided.)
Note on the chart, the calculated length from the nut to first fret. (The standard Martin measurement is 1.422; the short scale Martin is 1.394”.) Compare that length to the actual length from the nut to the middle of fret 1 on your guitar. If it is shorter than the calculated length, then you will need less nut compensation than if it is equal.
Measuring suggestion: It’s difficult to measure to the middle of a fret wire, so in order to get a good measurement from the nut to the middle of the first fret wire, use your caliper to make an “inside” measurement from the nut to the near edge of the first fret wire. Then measure the width of the fret wire, and add half of that to the first measurement. Return to top
In order to select which two frets to use when comparing pitch for adjusting the saddle(s), a player should carefully consider what portion of the fretboard needs the most accurate pitch, based on what portion she most often uses.
It's not usually a problem for acoustic guitars, because they usually are not played much above the neck joint. Compare pitch at the fret where the neck joins the body to the pitch at the 2nd or 3rd fret. That would be frets 12-2 for a classical guitar (it does not need to be an octave) and usually 14 - 2 for a steel string acoustic. That should give you good results up to about 4 or 5 frets above the high test fret. You need to use an accurate tuner - see: Comparing Two Notes - Cents-Off.
For very proficient acoustic players, needing accuracy well above the neck joint on an acoustic guitar (perhaps with a cutaway) it is first necessary that the fret levels above the neck joint are in good alignment with the rest of the neck, with no excessive falling off of the fretboard away from the strings. If that is a problem with a particular guitar, the options are to repair, upgrade, or just treat it as stretch tuning in the upper register (it's a feature, not a defect).
It is possible to use a higher set of test frets, such as 15-3, 16-4, or 17-5. Normally, the base strings are not played above the neck joint on an acoustic, so don't use a higher range on those strings that don't need it. The range does not have to be the same for all the strings.
Electric guitar players are much more likely to play higher up. For someone regularly playing all the way up to the 24th fret, I would recommend using 17-5. One luthier recommends, on his website, using 19 - 7, but I don't concur, because the highest frets are likely to go a bit sharp (see: The Clothesline Effect) and to try to correct them downward would end up causing pitches, starting a few frets down, to be flat. There is no easy remedy for the sharping, so I suggest just calling it stretch tuning in the higher register. (I have read that playing a high lead a bit sharp causes it to stand out from the rest of the band, which may be a good thing.)
The saddles on electric guitars are usually easy to adjust, so use your best choice of test frets and make the adjustments. It's a good idea to then take various sample measurements of the results, and if you could improve the intonation with a different set of test frets, go for it! Return to top
If someone wants to get serious about intonation work, or even, to be able to really check out an existing guitar, he must become familiar with measuring the cents-off error between two notes. For more information and techniques, please go see Tuners and Tuning. Return to top
I know that there are people who like higher action. These people are usually professionals or the more proficient amateurs. They are not bothered by the high action, and they like to feel the resistance wherever they play on the fretboard. I wouldn't try to change their minds, but they should be aware of the need for added compensation with the extra height.
However, I think there are way too many people who have just accepted what was handed to them, not really realizing what they could gain with a little nut slot filing. There are several advantages to having low nut action:
Most of us can play better with low action.
While higher action at the saddle, increases the sound volume on an acoustic guitar, higher action at the nut does not.
While higher action at the saddle reduces buzz from heavy play; higher than necessary action at the nut does not. (See the nut action check below.)
Higher action at the nut increases intonation error.
Higher action may hurt your fingers.
Higher action at the nut also increases the action height all the way up to the saddle. This is how it works: As soon as you play at the first fret, you have eliminated the extra nut height, and the action at the 13th fret will have been reduced by an amount equal to about half of the extra height at nut. This extra height is not useful, either at the nut or at the octave, and just causes more distance to have to move the string.
Most manufacturer's string height recommendations for nut action, I would consider to be excessive, especially on the bass strings.
Nut action check: While fretting at the first fret, check the fret-to-string height at the second fret. If the open-string height at the first fret exceeds the second fret string height by more than 2 thousandths, it is extra, wasted height. You can safely take the action down to as little as .0005" above the fret plane (as measured above).
As the years have gone by, and I play harder, I have had to raise the saddle to prevent buzz, but never the nut. Return to top
Many of you may have noticed more than expected resistance from the strings when fretting near the nut, especially at the first fret. I would like to use the clothesline as a somewhat euphemistic analogy of this intonation issue. I'm sure that you'll will prefer this to some trigonometric formulas.
Anyone who has ever hung clothes on a line has noticed that when you hang something near the middle of the line, the line sags a lot, yet when you hang something near the end, the line holds firm and resistant.
Stating this principle just a tiny bit more formally: The force needed to move a tight string a certain distance, perpendicular to the string, decreases as the applied force moves closer to the center of the string, and increases as the applied force moves towards either end of a string.
Now, the amount of force needed on a guitar string is the result of the the additional length the string that has to be stretched to in order to gain the required perpendicular movement. That stretch is maximized as you approach the ends of the strings, and that added tension affects guitar intonation.
There are several ways it relates to guitar intonation. Starting at the nut, and moving towards the saddle:
If the first fret action is low, there doesn't seem to be any problem with intonation from this affect, because the movement of the string to the fret top is so slight. With high action, however, there can be a significant effect, causing the first few frets, particularly the first fret, to play sharp compared to the intermediate frets, regardless of other intonation efforts, other than displacing frets. (This is distinct from sharping at the low frets due to lack of nut compensation).
As we move up the fretboard, toward the 12th fret, most of us are used to thinking that the amount of effort (and hence the amount of string stretch) is strictly proportional to the action height above each fret. The clothesline effect tells us that that this is not the case - that the amount of stretch is considerably less than previously expected, as we move up toward the 12th fret. So, in the middle range of the fretboard, we have to pull the string farther, but not harder.
This works out well for intonation up to the 12th fret, but beyond the 12th, the clothesline effect works against our intonation models, causing sharping at the high frets. This sharping is very slight until about the 17th fret (we're talking electric guitar and top acoustic lead players here).
I recently set up a Peavey electric with 23 frets, saddle intoned comparing fret 14 to fret 2, and compensated nut. Except for one area on the low E, I was a able to achieve intonation within 2 cents all over the fretboard up through the 17th. From the 18th to the 23rd, the pitch rose up to 4 or 5 cents sharp.
For a person who plays all the way up, we can intonate the saddle, at a little higher range, comparing, say, fret 17 to fret 5 or even fret 19 to fret 7 - I wouldn't advocate intonating any higher than that, because it would cause flatting at the mid frets. For any sharpness remaining at the high end, It seems best to me to just accept it as a bit of stretch tuning!. A little of sharpness on a lead gives it more "presence" - that may be a good thing (but I'm not the best judge of that).
Acoustic guitarists normally don't play that high, but acoustic guitar fretboards often recede away from the strings after the body joint at the 12th or 14th, aggravating the problem, and producing a similar sharping. I would suggest that luthiers be mindful of that when working on fretboards of finer quality acoustic guitars, and on electric guitars. Return to top
Using a zero fret on a guitar could minimize the need for nut compensation, because it eliminates the usual extra string height at the nut. (This assumes that the zero fret is not oversized.) This would eliminate the possibility of compensation on each individual string, but the zero fret could be moved forward of the calculated nut position, producing partial compensation. This may be a practical alternative to full nut compensation on, say, an electric guitar with light strings, and where intonation is pretty good to start with on most strings.
The same effect could be achieved by lowering the slots on the existing nut until the gap above the first fret is just a hair greater than the gap above the second fret while fretting at the first fret. For those who play hard, there may be a concern that wear on the slots over time might cause a buzz, necessitating replacement or shimming the nut. Return to top
Almost every guitar player does some sweetening, whether they know it or not. If the nut has not been compensated, then no chord will sound good without some adjusting.
Ideally, we think of sweetening as tweaking the tuning away from equal temperament, in order to produce more pleasing chords and natural intervals, at least in our favorite positions. We can do this by ear, or tune with a special tuner, according to various available tuning offset charts. But, in practice, I think that most of the tweaking is done because the intonation is bad to start with, the guitar never having been properly intonated to equal temperament!
Guitars with the Feiten System should be tuned to the Feiten patented offsets. Other sweetening charts are available with some of the Peterson tuners, including the Stroboflip, the Stroboclip, and Strobosoft. (The Peterson offsets are kept trade secret, but take effect when selected on the tuners.) Monte Allum offers his tuning offsets at a very low price, and I expect that other offset charts are available, as well. These "custom" offsets can be coded into the peterson tuners, or into the AP Tuner.
Ask yourself this question: "What will these offsets do if the guitar has not first been fully intonated, including nut compensation?". I suppose the best answer would be to use the computer saying: "garbage in, garbage out"! No, seriously, If you don't know How far "out" your guitar is, then how do know how much to adjust it, and what the effect will be?
In actuality, since most guitars lack nut compensation, most sweetening is done by ear, and is needed to make the guitar sound acceptable, and that usually means closer to equal temperament. Return to top
Acoustic guitars have skinny saddles – why? They are often insufficient for proper intonating of each string. Even when it is possible to intonate the 2nd and 6th string, there often is not enough material to properly round the string down toward the bridge pin. I would like to see all saddles 3/16" (.1875", .476mm) or more. 1/8" (.125", 3.175mm) should be the bare minimum, and that is inadequate for full accurate saddle compensation. (but where can you find a 3/16" saddle blank?).
For existing guitars, it is not absolutely necessary to re-route the bridge slot. Using a wide saddle blank, you could trim the bottom of the saddle to fit the slot, and form the top part as needed. (If you are extending very much toward the sound hole, be careful to insure that there is string pressure on the saddle in the opposite direction, in order to balance the load over the slot.)
Below: Anders Sterner compensated saddles, sternercapo.se
I saw some oversized blanks at: Guitar Parts and More.
So, would a wide saddle add too much mass to the bridge? (This might depend on each particular guitar.) If so, you could shape it to only extend where needed for intonation.
If you don’t play open strings at all, and you set saddle intonation comparing fret 12 or 14 to fret 2, then you could tune only to fretted notes and eliminate the need for nut compensation. This could be a good temporary measure, but you would probably want to do the nut compensation eventually – it’s easier to tune open strings than fretted ones. Return to top
For acoustic players who only play low on the neck, nut compensation is much more important than improving saddle compensation, and there are situations where readjustment of the saddle would be difficult or expensive. Adjustments you can't make without replacing the saddle could be skipped or postponed. First, check the amount of error when comparing fret 12 or 14 to fret 2, and decide whether you can live with it. Go ahead and compensate the nut, maybe leaving just a tiny bit of extra material if a string is flat at the 12th fret, and you plan to adjust the saddle later.
If you decide later to adjust the saddle, be consoled that it won’t effect your nut compensation very much. Nut compensation adjusts the first fret width, which is about 1.4“. The scale length is about 25.5”. A change at the saddle, would change the nut compensation in the proportion of 1.4” / 25.5” = .055. For example, if you later extended the saddle by .020”, that would only change the nut comp by -.0011”. Return to top
We all know that nut slots so tight that they bind the strings, will cause tuning problems, and can possibly even damage string windings. The tight slots cause the strings to hang on the nut slots. Therefore, when making adjustments, it is always good to be aware of keeping the nut slots smooth and even, just a little wider than the string width, and at a gentle downward angle at the leading edge of the nut (about 8 degrees). Return to top
I have not tried a lot of capos, so I can't tell you which is best. Instead, I ask you to focus on the relationship of capo features to improved intonation (while also considering practical usability and player preferences).
In order to minimize sharping, with or without nut compensation, the front edge of the capo should be placed right over the nut, or just behind. Also, the capo needs to fit the arch of the strings, and should not be tighter than necessary.
Here are some of the things to look for in a capo:
The Kyser capo is extremely popular: I previously used one for jamming, because they are so quick to apply, and the curvature fits the radius of my guitar fretboards. The problem is that they have a very strong non-adjustable spring (so they won't fail, even with a high action 12-string). The excess pressure detunes your guitar! To tone it down, I installed a 1 1/4” wide rubber band, made from a skinny bike tube. It works very well with that modification. The G7th Nashville Capo, for example, is also quick, but does not squeeze nearly so hard.
There are better capos available that are adjustable and don't interfere much with playing. If done properly, the capo will not noticeable sharp the notes.
Besides the right pressure, it's important that the arch of the capo fits the radius of your fretboard (actually the arch of the tops of the strings will be a little more shallow than that of the fretboard).
I have a Paige capo. It works really well for practice and recording, but it is slow to change in a jam, I was surprised that it was arched too much for my guitars. All was well, however, when I discovered that I could easily adjust it to fit perfectly! With the rubber removed, I compressed it with a vice, and then finished it off with a small hammer. The capo now works perfectly with a very slight pressure.
There are, of course, other good capos, similar to the Paige, such as the John Pearse, that many people like. These capos slide back over the nut when out of use, but some guitar necks do not allow for that, so try it on before you buy. And, of course there are many other styles that work well, according to your needs and preference.
Though I rarely bend a note, I'm aware that capos can be a problem with bends or tremolo bars, so I tested it. With the very light pressure of the Paige, the strings will slide easily, and return to in-tune.
Recently, I decided I wanted a capo with combined advantages: the speed of the kyser, and the accuracy of the Paige. My solution was the Planet Waves NS Tri-Action capo! It worked great, but like the Paige, It had a little too much curvature. It couldn't be fixed by bending, so I applied a little automotive belt dressing to the rubber, and carefully filed it to the precise arch that I needed. (Incidently, this method can be used to restore worn capos.). After reducing the arch, I was able to reduce the tension to just above the lowest setting, giving me just what I wanted in a capo!
With compensated nut and saddle, and with a carefully placed capo, even barred, and other closed chords, will play in very good tune (that includes being in tune with other instruments.)
However, when an "open" chord is played with capo, then the “open” strings could possibly play a very tiny bit flat, though probably not noticeable. That is because they will not have the increased tension from being fretted all the way.
Overall, playing with a capo is hugely improved with nut and saddle compensation. Exact? Maybe not, but greatly improved! Return to top
Fanned fret guitars have increasingly longer scale lengths from the first string through the sixth. The scales can be laid out by drawing the first string scale, and then drawing the sixth string scale. The other four scales are on a line between the first string scale and the sixth string scale at each fret. So, the frets are straight, but are placed at a slant which progressively shifts more counterclockwise, as you move up the scale. Each string scale is fretted in the same proportional relationships as on a standard fretboard – just different lengths.
The lower the pitch of a string, the more it needs length, in order to produce a rich tone; that’s why bases have longer strings than a guitar, and guitars have longer strings than mandolins.
The main advantage of the fanned fret designs is that they provide that extra length for the base strings, producing improvements in both tonality and harmonicity (better refinement of overtones on the bases) while maintaining the desired scale length of the trebles. Although these results are significant, the length differences are small compared to those of a piano.
I have read forum posts saying, to the effect, that you cannot fully intone a guitar unless it has fanned frets. This is very misleading! The principles of intonation are the same for fanned fretted guitars as they are for other guitars. However, the bass strings being longer, may require a little less compensation.
Fanned fret guitars are said to be more ergonomic, but there is disagreement about that. Maybe it is different depending on playing style. I can’t say, because I’ve never tried one.
Stretch tuning means progressively making higher notes more sharp. Pianos are usually stretch-tuned. That is because, at the highest octaves, there are problems with the overtones mis-matching the notes on the octaves below, and at the lower octaves, there is a problem mis-matching the overtones with the lowest octaves. Correction for these problems results in an S-shaped curve of intonation, with the higher notes being raised even higher, and the lowest being lowered even more. On a piano, each note is intoned separately. Because of longer string length, concert grand pianos need the least correction from equal temperament.
On a guitar, however, notes cannot be intoned separately, so, in order for a particular note to be the same on multiple strings, for stretch tuning, it would be necessary to stick with a constant rate of pitch increase, the same on all strings. The guitar does not have nearly the range of the piano, so there is really no need for stretch tuning. The piano plays 8 octaves, while the guitar plays only 4.
Stretch Tuning should not be tied to nut compensation, but it could be used in conjunction with nut compensation. As I understand the Buzz Feiton System, it does not seem to use overall stretch tuning, but uses stretch on some strings, perhaps to effect particular harmony tweaks.
Still want to try it? To effect stretch tuning, the strings are shortened at the saddle so that the notes from low to high are progressively sharpened - by how much? Pianos stretch about 2 1/2 cents per octave, on the lowest and highest octaves. It would depend on preference – try it and see if it sounds nice to you (usually not difficult on an electric guitar, unless you reach the saddle adjustment limits). Remember to consider what other instruments you will be playing with, and how they will be tuned. Oh, and you’ll need a special tuning procedure!
As an acoustic guitarist, stretch tuning holds no appeal for me, but if I were playing high, soaring solos on an electric guitar, I might possibly think differently.