Fretscale errors

When you have a guitar with a bad fret scale, whether it's due to randomized error, or a non-unified scale, it can seem like the guitar will not play or stay in tune. 

Guitarsmith Richard Stanley says: I found out in the mid 70s that fret scale problems were a significant issue. By the way, during my first few years in this business, if someone had told me: "You know what? You're going to find that a significant amount of your bushiness is going to be replacing fingerboards — because there are problems in a lot of those fret scales out there" I would not have believed them. But players just kept coming around saying: "I've got a problem with this" and it still was the last thing I would have thought to look for. It took quite a while before I got around to considering the possibility that there was a problem with where the frets were located.

But I found that players were pointing out areas where they heard intonation issues that had errors of just 1 percent of the distance between adjacent frets. It is important to remember that I didn’t go looking for this issue. It took many players expressing dissatisfaction with the intonation characteristic of their instruments over a period of several years to attract my attention. Most players who were contending with faulty scales think they are dealing with a pitch stability problem such as string slippage, sloppy tuners, rather than intonation. 

I finally got to the point where I'd heard from enough people that I thought there's really something wrong here.  When I started looking at the fret scale issue, I found two distinctly separate deviations:

  • Randomized errors over a period of time where no two were the same
  • Non-unified theoretical string lengths, which appear to be intentional modifications of the fret scale. These take the form of scales having more than one theoretical string length within the fret scale, some with as many as five different theoretical string lengths incorporated in one single fingerboard.

Some of the many fingerboards that Stanley has replaced over the years.

Randomized Error

This is what randomized errors look like:

Here's a particularly bad scale on a US Guitar Kit instrument (above). The fret scale is all over the place. The fingerboard nut is flat by 33/1000 of an inch, and some of frets are off in the 20s. A few are right on indicated by star), but 7-11 are sharp, then 13-18 are flat again and so on.

Prior to about the late 70s, Fender had a fairly recognizable amount of random error. Unified scales, but random errors. You'd rarely see two Fenders that would survey exactly the same. Then, also in the 70s came the Japanese Squire production . . . spot on. Dead on. Then in about 1980, bingo, the American Fender production all of the sudden snapped into place. So they must have re-tooled at that point, and got a more precise operation. Now, Fender stuff seems to be looking a little shakier again. And I presume that it's time for them to re-tool again. They're still pretty good and generally acceptable, but I'm not seeing spot-on surveys as often. I never got a look at enough examples from the first five years of Fender production in the early 50s to see what the story was then. Most were from the 60s and 70s, when that gear would have been worn out, or maybe it was a more primitive type of setup that didn't provide as much precision. 

Non-Unified Scale Lengths

Now look here:

Here are surveys for a 60s SG and a practically new SG. Look at how similar they are. They aren't exactly the same, but they're close. This is how all Gibsons tend to look (to this day). The first four frets are typically 24.5, and in the upper end, the frets are somewhere around 24.75, sometimes running out as high as 24.85.

Here's the original board from Dave's 1954 Les Paul.

You can see from the survey that from frets one to five, the fret scale is 24.5, from frets five to 12, the scale is 24.6, from frets 12 to 16, it's 24.7, and (finally) from frets 16 to 22 the scale is 24.75.

24.75 is what we think all Gibson fret scales are supposed to be. But on this board it was only 24.75 from 16 to 22. And the rest of the board was segmented into four different, non-unified scale lengths.  This type of deviation is readily distinguishable from random error by the fact that many, many different examples of this form of error are exactly the same, repeated time and again. I refer to this type of segmented scale as a multi-theoretical string length scale or just multi scale

So what you'll see is one group of frets will agree with 24.75, and another group correlates with 24.6 and so forth. And where a random scale is all over the place, this is purposeful. There's no two ways about it. You can't look at guitars made over a 40-50 year period and continue to see this pattern without realizing that they meant  to do it. 

The practical effect is that the string lengths cannot be set up so that all areas of the neck play reasonably in-tune. If you set the bridge saddles for the true octave, you find that the 19th fret is sharp. Or if you set it up off the 19th fret, you'll find that the octave is flat. This was obvious to me at the outset even before I had the templates that let me quantify the problem.

Editors Note

The preceding discussion on fret scales is designed to promote awareness of issues that may or may not affect you. It is important to note and remember that rock music history has been made, recorded with, and played live, with thousands of guitars that most likely had "imperfect" fret scales. Millions of musicians and fans never noticed or heard a problem, much less complained about it. Probably because guitarists who hear tuning/intonation problems severe enough to bother them don't typically dwell on the reasons why it is happening. They usually just get rid of the particular guitar and replace it with one where such issues are imperceptible. This is a perfectly valid solution.

The key thing to take away from this discussion and Richard and I can't emphasize it enough is this: If you don't hear a problem, you don't have a problem.