Official Luthiers Forum!

I was wondering if there is any consensus on the value of a compensated acoustic saddle. I get that an instrument is either in tune or its not, and so in that regard obviously a compensated saddle is better than one that is not, but I wonder how much benefit are you really gaining with a compensated saddle since there is always an element of compromise with a fretted instrument.

A few years ago, I had my D-18 set up with a compensated saddle and it's intonation is pretty good, give or take a few cents. I've got a Guild D-50 in my shop right now that looks like it was set up by the same luthier with a similar looking saddle to the one in my guitar, and the intonation is much more inconsistent than on my guitar. I've got another guitar here as well, it's a Blueridge 000-18 copy, I just replaced the saddle with a bone saddle that is not compensated and the intonation seems to be similar to the Guild.

This brings up two other questions. First, other than the obvious ear test, that is the guitar sounds really out of tune when you play it up the neck what do most people consider to be an allowable tolerance for intonation? I always make sure electrics with individual saddles leave the shop intonated, but with an acoustic the matter is a tad more labor intensive. Which leads me into the second question, is compensating an acoustic saddle simply a matter of trial and error, incrementally filing each spot on the saddle until you get the sweet spot? Or do people just have a specific contour they use every time, and hope that it is a better compromise than a non compensated saddle?

The real test on this is the client - do THEY care if it's slightly out of tune?

I think it makes a diff, and it is a real pain to produce such a saddle. The sad fact about that is this: If the client changes string gauges, or even brands of strings - the intonation will change. They don't want to hear that.

I have an elaborate process for making a compensated saddle that involves routing for a slightly wider saddle blank, gluing up said blank or making one from a large piece of bone, and then using a short piece of .010 guitar string as a temporary "peak" under the string that I can move back and forth to determine the intonation, marking the position (using a sensitive tuning device), and repeating for all 6 (or 12) strings. Of course, you want the same gauge of strings the client intends to use, with the saddle radiused to fit the fingerboard, and the action set as close as possible to where the client wants to play the guitar. It's quite a juggling act, but it can done if you are careful to measure frequently, and patient enough to follow the process to the end. Finally - you will charge accordingly, like 3 times as much as you would for making a normal saddle.

_________________
"Act your age, not your shoe size" - Prince

I always compensate acoustic saddles mostly because I can easily do it. Saddle work with files is very enjoyable to me especially getting a high polish on our nut buffer toward the end.

Humans generally can hear differences of 4 cents or more, some of us better than that. I set intonation when I can (meaning if saddle travel or position permits it.... and it doesn't always happen this way....) to as close as I can get to right-on. "Right-on" is of course a very scientific measurement..... I use Peterson Strobo-Soft on my iPhone.

Mind you check intonation and then press harder or less hard and watch what happens to your readings. Intonation is also a function of how far an individual player stretches strings by finger presure which is also why low nut slots are critical to quality in a set-up.

Anyway when setting or checking we want 2 cents or less before I move on the next one to do. I shoot for the same light finger pressure and being exactly on the number but again all bets are off with different pressures on the fret board. Sometimes you have to take an average on instruments with one saddle for two strings - Teles, etc.

It's been said that without checking a thing simply making saddles with the high e in front of the b, the G in front of the b, d, a and the low e where the b is and this will be an improvement for most acoustics with a wound g. It's been my observation that this is true and this is also why Martin and others sell's pre-compensated saddles just as described for drop in use on Martins.

The only time I don't compensate a new saddle on an acoustic is if it's historically correct not too AND the saddle is too thin for compensation to not make a difference.

As a player I want my acoustic intonated as close as it can get, especially since I often play in an orchestra setting. In my shop, intonating the guitar the best that it can be is a standard part of the setup. I usually charge for two sets of strings because the first set generally gets abused during the setup process and I want fresh strings on the guitar when it goes out. I pretty much follow the same procedure that Chris does using a small cutoff piece of string, marking the location for each string then cutting the saddle to fit. I always warn my clients that changing string gauges or brand can cause problems with the setup.

Of course, sometimes it's not practical, cost effective(cheap student acoustics), or even possible to intonate some guitars.

_________________
Steve Smith
"Music is what feelings sound like"

Connor, if you haven't been already, I highly recommend going to the Ann Arbor web-site.
Dave Collins made a few videos that must be the most concise intonation overview available.
Just keep your cursor on the pause button and turn on the information fire hose.

http://www.annarborguitars.com/styled-4/index.html

Lets do a flow chart.

You've measured the guitar when you got it and know how much the intonation is off. Is it acceptable to your customer? To you? I read somewhere that most of us can tell 5 cents difference, often not less. Most electronic tuners can indicate 2 cents, a strobe tuner 0.

Second, are you planning to make a new saddle? Why?

Third, have you measured the pre-compensation? Is the saddle slot in the right location (scale length plus a bit on the high E, scale length plus three bits on the low). If the slot is in the wrong place should you fix it (valuable vintage guitar, maybe not, 1970's Martin, probably)? Are you equipped to fix it?

Is there a reason why ordinary compensation might not work? This might include split saddle, compensated nut, multi scale, weird strings or tuning, slide guitar, nylon strings, old saddle is some exotic material or is glued in....

Are you going to lower the action? That usually improves intonation.

If you are going to make a new saddle anyway and the old one wasn't dead on, why not compensate it some more? Do you want to go thru all the work that Chris describes with the little piece of wire (a B string works well)? I've done it on two twelve strings - it is a lot of work that you may not be able to charge your customer for.

Final decision - if you are going to make a new saddle and the ordinary compensation (high E forward, B back, G forward, low E back) looks like an improvement over a straight saddle it is very little more work to make. If it is a classical you might consider compensating the G to the back.

Bottom line, unless there is a compelling reason not to, I do the standard compensation on every saddle that I make.