Hi, thanks everyone for your help and advice in the past. I searched for discussions on this topic and still thought I might get some new insights if I created my own post.
I've done a few refrets and have been trying to understand how to get my guitars set up better in general. Much of my information has come from Dan Erlewine's videos and his book Guitar Player Repair Guide.
I heard Dan say repeatedly that he prefers a "straight neck" and that he's not a big fan of relief, so I simply aimed for dead-flat necks on my guitars. I ignored the idea of any relief, however slight.
Now I'm thinking that a little relief may be an important missing element in my set ups. I found a discussion of acoustic guitar set up which got me thinking. It is from someone who appears to be a hobbyist, however it seems well-thought-out to me---http://thbecker.net/guitar_playing/guitars_and_setup/setup_page_01..... I want you guys' opinions, though because I'm still trying to wrap my mind around how all the factors contribute to the overall set up and visualize what's going on.
Becker's method is basically setting in the desired amount of neck relief first (he recommends .010"), then setting the saddle height with a capo on the first fret, and then filing the nut slots down.
Erlewine recommends cutting nut slots with the neck adjusted dead-flat, so that relief can be added if desired, as opposed to cutting them with relief, and running the risk of someone coming along later and straightening the neck and finding lots of buzzing. So, a dead-flat neck requires the nut slots to be slightly higher to achieve the same action at the first fret, right? This seems logical. And a relieved neck with "X" clearance at the first fret will have slightly higher action at the 2nd, 3rd, 4th frets, etc. than a straight neck with the same action at the first fret. Is this correct?
As I imagine the path of a vibrating string plucked open or at the 1st or 2nd fret, the idea of relief makes perfect sense to me, as the fretboard could theoretically be made to follow as closely as possible just outside the "cigar-shaped" area (term I read on the forum somewhere---this would be a long, thin, pointy-tipped cigar) through which the string vibrates. The frets would follow the underside of the "cigar" beginning at the tip near the nut, and when reaching the midpoint of the cigar, would no longer follow back up, but remain level to the end of the fingerboard (or fall slightly away at the tongue).
I still am not sure how to visualize how a relieved neck relates to the path of a vibrating string as you move up to the 4th, 5th frets, etc. I do see that on most guitars, the action is lowest at the nut, with the strings rising away from the frets slightly as they move toward the bridge. It makes sense to me that relief mainly serves the purpose of reducing buzzing in the lower frets, as the higher frets are less susceptible since they must be pressed further when fretting, with the vibrating portion at a safe distance from the higher frets. So, in the "flatten-out area" as Dan calls it (frets 9 or so and up), I understand that buzzing would be less likely, but what about the area between the low frets and the beginning of the flatten-out area?
Is there a general consensus about relief vs flat necks? Are certain tradeoffs inevitably involved, etc.? If I imagine the "cigar" shape in relation to a flat neck, it seems that a higher saddle would be required than when the neck has relief, as a cigar whose tip is at the nut would have to be upturned away from the fingerboard (like if you had a pointy-tipped cigar, or a toothpick, etc. resting on a surface and you press one end down toward the surface, the other end is lifted up). Is this correct? Can I assume that Dan plays with a very light touch? Otherwise I'm not sure exactly why he prefers no relief. I also play pretty lightly in general (kind of James Taylor-style fingerpicking and light strumming, no single-note solos up the neck on acoustic--I'd also like to leave some room for strumming with a pick, though I'm never too aggressive), so I figured I'd go for a very small amount of relief to get my ideal playability.
I would love to hear anyone's thoughts on all this! Sorry it was so long---hopefully someone reads it.
Ahhhh... relief. Now there's a topic. And it's one on which the experts do not agree.
One highly respected luthier says that he prefers to get the neck truly flat, with just a little relief. Umm - to me flat is flat, relief is not. And while some prefer a truly flat neck, others will insist that a little relief is helpful and can lead to overall lower action if the frets are really levelled and all else is equal. Myself, I prefer a little relief to help out a bit in the area of the first few frets. And it seems you have a pretty good grasp of how that geometry works. Just to add a bit to that understanding, it may be helpful to realize that the largest excursion of the string is when the string is played open (not fretted). You can easily see this for yourself. Pluck an open string and look at the mid point on the string length and see how far the string displaces in it's vibration. Then fret the same string at the last fret and pluck it with the same force. You will see that the string excursion is much less. So you might draw the conclusion that the closer to the nut you are fretting, the more the string needs clearance from the frets. And since the strings are closer to the frets as you get closer to the nut, they may well benefit from a little relief.
String height at the nut is another matter. This is not an arbitrary height - there is one height that is right assuming the rest of the geometry has been set properly. I've heard plenty of ways of determining what is "right", and many of those methods yield acceptable results. Here is the consideration (assuming all the other geometry is indeed correct). If you fret at the first fret, there is a specific clearance between the string and the second fret. That is the ideal clearance you want to have at the first fret when the string is open. Any less and you are more likely to get fret buzz on the open string than when fretted at the fist fret. Any more and you are providing more clearance than is necessary compared to fretting at the first. The consequence of any excess of string height on the open string will be that it will unnecessarily stiffer to fret at the first couple of frets, and the strings will be squeezed slightly sharp.
But there are two other considerations we need to address when cutting the nut slots to the right height. First is that the nut slots will wear down over time from tuning which drags the string through the nut slot, abrading it down somewhat. So a tiny bit of extra height may be useful.
Secondly is consideration for the condition known as back buzz. This is where the string vibrates against the frets between where you are fretting and the nut. There should be a tiny bit of clearance between the string and the frets "behind" where you are playing to avoid this problem. And that clearance is provided by a bit of relief and enough height at the nut.
Thanks Richard. This helps a lot--I will reread it several times to try to absorb everything you are saying. I really appreciate you taking the time to explain.
I did experience back buzz in the past when setting up guitars with a straight neck, and at the time I wasn't sure how to deal with it, but it's becoming clearer now that a little relief should take care of it.
"Erlewine recommends cutting nut slots with the neck adjusted dead-flat, so that relief can be added if desired, as opposed to cutting them with relief, and running the risk of someone coming along later and straightening the neck and finding lots of buzzing. So, a dead-flat neck requires the nut slots to be slightly higher to achieve the same action at the first fret, right? This seems logical. And a relieved neck with "X" clearance at the first fret will have slightly higher action at the 2nd, 3rd, 4th frets, etc. than a straight neck with the same action at the first fret. Is this correct?"
We cut the nut slots with the neck relief pretty flat, not necessarily absolutely flat, exactly for the reason of not wanting to cut the slots too low.
Here is where I think there is some misunderstanding here. True action is a function of three things generally speaking..., nut slot depth, relief, and saddle height. Where we go astray in understanding set-ups is thinking that we set action with the nut slots.
True slot depth is one of the things that play into action but nut slot depth should be cut independent of any desired action. Nut slot depth is a variable and as such can be set very well independent of the other variables if we understand the relationships and dependencies and then this variable is eliminated and action becomes a function of the two remaining variables relief and saddle height.
I'm not sure that Dan means absolutely flat and there is some important context there too. Where a very flat neck may work well for some players set-up with higher action a heavy hitter with very low action is not going to be able to get away with no relief.
A great set-up factors in the player to a very large degree and cannot be done without considering how someone plays and even what they play at times too. We tweak relief, slot depth, action height all to match the application/requirement/player. Generic set-ups can work well too but when you are getting into the nits of super flat necks the player becomes a variable and consideration too.
As such independent of nut slot depth which can be set and then eliminated as a variable the set-up that works best for you will also be dependent on your playing style, what you play and even string gauges too since less massive strings can lash out more and create fret buzz because they have less mass and inertia.
Our books and guides can explore these concepts and answer some questions but ultimately if a dead flat neck is right for you will also be a function of You! :) With this said a little relief is likely a good thing for you and most of us humans bags of mostly water (Star Trek reference..).
And yes someone with a dead straight set-up likely plays with a light touch. Kind of hard to imagine for Dan when back in the day he hung out with and played with Iggy Pop here in Ann Arbor...:)
"A great set-up factors in the player to a very large degree and cannot be done without considering how someone plays and even what they play at times too. We tweak relief, slot depth, action height all to match the application/requirement/player."
Expertly put Hesh and aside from the ability to accomplish the proper tweaking of the nut, saddle(s) & neck relief, this is, to me, the most important factor to understand in a great set-up. KNOW your player. KNOW their needs.
One more thing that I wanted to mention.
When seeking minimal relief or no relief the quality of the fret plane, how the strings see the frets since strings are natural straight edges, becomes even more critical. Substandard fret work is not going to permit us to push a set-up to the limits and be happy with it.
Decent fret work can produce results with a variance of less than .0005" and although many would not see why this level of precision matters the grease ball, weekend wedding player who wants action of 2/64" and 3/64" measured at the 12th will require excellent fret work and very precise relief.
The more the set-up requirements push the spec envelope the more the fret plane becomes important as to it's quality and levelness. With this said the forth variable to a decent set-up might be the quality of the fret plane and that's what I wanted to add. Just takes me way too many words to add anything sometimes....:)
For the frequent forum fliers ever notice that we never get these questions from classical players? Perhaps the reason why is with action of 5/7 or higher the quality of the fret plane level matters less, relief too. But for the shredders out here it's a different story....
Nice work Hesh, as usual. My add is that the modern applications of compounding and drop away past the 15th fret (or thereabouts) has changed the way we set up and my latest thougths involve the permutations of three points in space: the bridge saddles (and their radius), the nut (and I don't fixate on nut slot height as the change in action height has very little to do with the angular reationship of the nut to the first couple of frets) and where both these points relate to the fret board up past the 12th fret.
The aim is to get the elliptical path of the cowboy chord strings taken care of by the relief (I don't do "completely flat", even for shredders) and as the angular realationship of the string to the fret increases as we go further up the board, thereby providing clearance for the trailing frets, I go flat and then into a slight drop away to provide full volume at full bend (no choking). How much compounding that is dialed into the frets to assist clean bending is a function of the players style and/or requirements and also string gauge (a major contributing factor to set ups as I see it).
However, a rather exceptional 10 year old virtuoso metal player came in the other day to get his guitars set up for Australia's Got Talent and totally floored me with his demand to be able to pull a couple steps on his Floyd while bending a couple of steps up at the 20-24th ( it was a lot of steps anyway!) without bottoming out. I had set the guitars up to shred and, yep, they bottomed out because the Floyd was going over center and lowering the strings as he pulled way back to get the thing screaming. I learned a very good lesson from a 10 year old who looked at me and said: I need it to do this. Lots of drop away on these guitars now but only up where this technique is applied. Learn something about set-ups every day.
Great subject, I would love to have a Plek to do all the donkey work and test and evaluation.
When someone talks to me about a "dead flat neck", my first question is usually; "Are we talking about a neck under tension or not under tension?
I'm certainly no expert but I haven't seen very many "dead flat necks" with strings tuned to pitch.
A lot of the stuff I've worked on was old enough or cheap enough that there is no adjustment whatsoever even if it happens to have some sort of reinforcement. Honestly, I'm not good enough at resets to do more than "take it into account" when I do them and I've had to redo more than one because this "technique" wasn't as effective as I hoped. As a result, I've come to like a stiff neck with a bit of flex but not too much. Sometimes I end up doing an "adjustment" to the neck with a switch in string gauge.
"Erlewine recommends cutting nut slots with the neck adjusted dead-flat, so that relief can be added if desired, as opposed to cutting them with relief, and running the risk of someone coming along later and straightening the neck and finding lots of buzzing."
I don't bother with any relief adjustment in order to set depth at the nut slots. I eliminate all factors down stream by fretting the string slot I'm working on between the 3rd and 4rh frets. I then measure the distance from the top of the first fret to the underside of the string with individual feeler gauges that have been removed from the set so I can feel better with them. I have documented my results over time and have settled on .005" at the treble E and add .001" to that for each larger string, arriving at .010" at the bass E slot. I suppose that this method is actually simulating a flat neck but takes me one less step to get there.
Totally in agreement that assessment of the player's style is the first step in a set up. Any special needs should be aired out prior to any work. If they don't know any better, they get my default values, a decent low action geared for Cowboy chording. However, I am often amazed at the action issues some really great players will put up with. I run into this more than I do players with a definite action shopping list.
Ned, my take would be strings on and to pitch, if the discussion is about action.
Why would you graduate the depth of the nut slots? Consider that when you fret at the first, there is a finite amount of clearance over the second fret. What is the purpose of having more clearance than that over the first fret when the string is open?
For the same reason that you have a tapered saddle, the larger the string, the more wobble room you need when it is vibrating.
I measure string heights that are lower than my values all the time that are not buzzing. The numbers I use are a safe place to set them for a new nut set up and leaves a bit of space for string wear. If you put all the strings at the same height, you would set the treble strings higher than they have to be, in order for the Low E and A strings to be set where they won't buzz when played open. The reverse as well. Try setting all of your strings at .005", the way I measure them and see what happens.
So do you have more clearance on the fatter strings over the second fret when fretting the first? Even with higher saddles on the larger strings, the effect at the first fret will be far less than you are suggesting. I do use about 5thou string-to-string at the saddle, but that does not translate to 5 thou at the nut. I fret at the first, measure the gap at the second, then reproduce that clearance at the nut for the open string -for each string. It strikes me that having 20 thou more clearance on a B string than the G string is excessive. And any advantage there may be is negated as soon as you fret that string.
Having a bit of a hard time following you Richard.
If you fret at one and measure at two, then that measurement would be affected by both relief and saddle height. The way I do it, I fret between 3 and 4 and measure from the top of the first fret to the bottom of the string. That eliminates variables such as relief and saddle height.
"I do use about 5thou string-to-string at the saddle, but that does not translate to 5 thou at the nut."
Don't understand what you mean.
"It strikes me that having 20 thou more clearance on a B string than the G string is excessive."
If you re-read what I said, .005" at the treble E and add .001" to that for each larger string, arriving at .010" at the bass E slot, the difference in the nut slot from B to G is .001". I'm not sure how you arrived at .020". The amount of change from E to E is .005", less than the width of a typical hair.