I Think My Frets Are In The Wrong Place....

[Crusader]

Kingfisher that experiment you are doing is just what I have always wanted to do

...Is that a rusty old door hinge you're using as a tailpiece on the left? ha ha!

Well I have just been crunching some numbers and guess what?

Using 1.058874 instead of 1.059463094 gives me the same result in a single formula as using the 3 different scales method I was using

And to convert it to the type of number Kingfisher is using you calculate as follows

1/(1-1/1.058874) = 17.9854265

But you can round it to 17.985

Calculating with 1.058874 or 17.985 in combination with 628.65 (24 3/4" not 24 9/16!!!)

will give you the measurements I got off the LP except for 3 differences of 0.1mm

I use excel and it goes like this to get distance from nut to fret

=SUM(1.058874-1.058874/628.65^1) = first fret 35mm

=SUM(1.058874-1.058874/628.65^2) = second 68mm

=SUM(1.058874-1.058874/628.65^3) = third 99.1mm

However what I do is have the big numbers in cells 27 and 30 so I can change them so the actual formula looks like this

=SUM(G27-G27/G30^14) etc

While inputing the formulas I use 'x' and 'y' then Copy and Paste then use "Find and Replace"

otherwise G27 ends up being G49 at the 22nd fret

Getting the numbers 1 to 22 is still laborious though. You have to press F2 and go through each cell and rename them, but once you're done you can copy and paste the whole lot into another section and change the scale length and calculating number

ooohh I spend too much time doing this sort of junk

Hope that helps, cheers!

[Crusader]

I just thought I would chuck in a couple more comments

I found in my notes from ten years ago the number 1.058784784 when trying to figure out the SG scale

It matches fairly well so I don't know why but I abandoned the single formula concept

Using the 3 different scales method gets the same results but its so laborious

Further tinkering with numbers has revealed this

1.0588 matches the SG but the 5th fret is still 0.5mm out - that has probably got a lot to do with why I abandoned it

I re-measured the LP a bit more today and am convinced that several frets are 'exact' in inches (I have always measured frets in both metric and Imperial)

For example the 22nd fret is 17 23/32" which converts to 450.0563mm

Using these exact imperial measurements I came up with 1.0588705

Whichever number Gibson use, this is surely the method to get these fret placements, not the three different standard scales

Even though I spent countless hours measuing and calculating over the years I am so glad to finally have this sussed. I call it the Kingfisher method lol!

WesV how do you get images of Excel on here? do you copy & paste into MSPaint or something?

Kingfisher I was thinking with those experiments you could try doing it with super-low action

That way you could find out how much compensation you need just due to thicker strings

Edited April 1, 2010 by Crusader

[WezV]

i used the snipping tool on windows 7, but you can just press print screen and paste it into paint before trimming it down

[kingfisher]

What do you mean "rusty old door hinge"

The tailpiece is a repurposed ferrous artefact covered with a copper/zinc alloy, originally designed for variation of angular placement relative to a entrance frame.

When I get a minute (too busy worrying about Rooney's ankle at the moment), I will redo the experiment with different string gauges. I purposely used the most commonly used top E gauge, and set the "action" to what I would regard as pretty damned good at the 12 fret.

We already know that thicker strings behave differently, - your bridge is never straight, - but on the other hand when playing in the oxygen thin atmosphere of the top of the neck, we tend to be on the top three strings - (ever noticed how the tuning up there is very flat on the bottom three strings).

Gibson may have optimised the number for thicker strings - hence our disparity.

One thing I am convinced of - if you use the mathematically pure 17.818 (17.817 if you are Jewish) calculation - your instrument will NOT be in tune

Sadly I am afraid this may be a leap too far for some (even with the Gibson connection) and wrongly fretted guitars will still be made.

"The SG is 0.5 mm out" - sheet, I do my fret slots by hand with a saw - I dream of being 0.5 mm out.

[WezV]

One thing I am convinced of - if you use the mathematically pure 17.818 (17.817 if you are Jewish) calculation - your instrument will NOT be in tune

Sadly I am afraid this may be a leap too far for some (even with the Gibson connection) and wrongly fretted guitars will still be made.

"The SG is 0.5 mm out" - sheet, I do my fret slots by hand with a saw - I dream of being 0.5 mm out.

we all know the problems inherent in the guitar, and a few solutions that help improve it.

might be worth taking your claims back a notch till you have actually got a whole guitar fretboard working with your system

dont forget you still have to see if it actually gives improved intonation on all frets and strings.... and this is where i think you will start going a bit crazy

this is the claims from the earvana site, showing how 'imperfect' a normal guitar is compared to one with an earvana nut

shame they only give notes up to the 12th fret - be interesting to see the rest, my guess is they dont show it because there would be some more red notes

this is where i get confused about what you aim to achieve ??? simply having less bridge compensation doesnt seem like that big a deal unless you are working on improved intonation. And any existing system that gives improved intonation involves individual differences for each string, be it bridge compensation, nut compensation, the true-temperament frets or a combination of all 3. these are things that take us away from the single number maths involved in fretting

simply put - i dont see how your system will be more 'in tune' or be more 'correctly fretted' compared to a normal one, although i can see how it might need less bridge or nut compensation to get it 'in tune'

also, iirc the number used for the maths is 17.817152. so it wouldnt be 'jewish' to round it to 17.817, just more accurate

[kingfisher]

Using 17.818 on a 650 mm scale I get 324.989 = 325 - 0.011

Using 17.817 on a 650 mm scale I get 325.002 = 325 + 0.002

I stand corrected on that point.

John.

As for the Earvana stuff, that is irrelevant to the argument. My theory ONLY relates to the octaves. Earvana assumes perfect octaves and is dealing with the stuff inbetween.

!7.817(ish) will not give you perfect octaves, so presumably the Earvana stuff is a non starter without you sort that first.

Edited April 1, 2010 by kingfisher

[WezV]

try it with 17.817152

650mm scale from nut to 12th fret is 325.000022714102mm

still maths and impossible to cut numbers

[Prostheta]

Bleh, I wait for the day when a pill is invented that makes my dissonant out-of-time playing music for the educated ear. Balls t'yer compensation lad.

[kingfisher]

Bleh, I wait for the day when a pill is invented that makes my dissonant out-of-time playing music for the educated ear. Balls t'yer compensation lad.

Good argument, well presented.

[Prostheta]

Bleh, I wait for the day when a pill is invented that makes my dissonant out-of-time playing music for the educated ear. Balls t'yer compensation lad.

Good argument, well presented.

I learnt everything I know from the televised House of Commoners.

[Crusader]

"The SG is 0.5 mm out" - sheet, I do my fret slots by hand with a saw - I dream of being 0.5 mm out

Laughing Out Loud I know what you mean

I hope all this isn't too much but -

The first pic is a page form my notes (check date top LHS) and below it is my measurements off the SG

I was trying all sorts of ways to figure out the SG frets are done

Over on the right is a long number and a list below it. Either side of the list is how much it differs from the measurements I took

.1 .04 .1 .11 .48 etc etc

The numbers weren't accurate enough for me...***?

In this next pic I show how the SG calculates via the normal formula vs "The Kingfisher" method (haha)

Note how the number is more than 18...huh? but it matches the SG measured spacings the best. The number that matches the LP is a bit different

Lastly I checked the intonation of the Strat Copy that I made

It has basically a Gibson fret spacing on the 6th string side and a Standard scale on the 1st string side

The frets are not exactly where I wanted them to be though (I am a lousy fret-worker)

Hope all this is helpful or at least interesting

cheers

Doug

Ps Wes, don't look at the numbers.....haha

[Prostheta]

I think it is possible that the Gibson "short 1st fret" combined with the compensation at the bridge is a simple step towards compensation from both ends of the fingerboard. Consider an Earvana nut or Buzz Feiten for example. These relieve the chore of compensation being applied purely by the bridge saddles by compensating the nut also. These seem like methods which are in fact more achievable by the lay builder. A slightly more complex method can be achieved by having a *proper* tuner and progressively adding shims to the string side of the nut as compensation.

BTW - did anyone know that when recording, Satriani actually uses Buzz Feiten and "standard" non-compensated nut instruments interchangably depending on the requirements of the instrument for the song? Sensible.

[WezV]

having a *proper* tuner

i did mean to mention this earlier, i have a peterson stroboflip and it works great... although some people dont like the interface

although i must admit i was a bit miffed recently when i saw you could get a peterson istrobosoft for the iphone/itouch for £5.99 !!!! 1/10th cent accuracy so, in theory, its as accurate as the £160 stroboflip

consider something industry standard like the boss tu-2 is +/- 3 cent

[Prostheta]

I like my Korg DT-10 for simplicity of the stompiness, but could do with adding a better tuner to my rack.

[kingfisher]

Thanks for that Doug - very interesting.

[kingfisher]

And as regards tuners, in an ideal world I would like one with a readout of the actual cycles per second, but I don't know if anyone makes (an affordable) one.

Also, maybe we are a bit at cross purposes here. Are the Earvana and Feiten systems more designed to give a better compromise for chords? (we all know if you tune to a perfect sounding open E chord at the bottom of the neck, and then play a C chord, again in the first position, the C chord sounds way out).

This would explain the Satriani thing, depends on what you are playing - if you were only playing E shapes up and down the neck -who cares about C shapes?

The "Kingfisher method" (if I may humbly use that term) is more about putting the frets in a better place for up and down the board on a single string.

Now, moving the frets about as I propose will certainly affect the chord thing - but how I don't know. - is this E/C thing better or worse on a Gibson or a Fender? - how should you tune the guitar (fifth fret, octaves or WHY).

It is my untested and unsubstantiated belief, with no evidence whatsoever, that the chord thing will be affected in a positive, good way.

[brian d]

FWIW, the MIMF forum library has a fret position calculator called "fretcomp" by Chris Franklin which "computes the pitch errors caused by fretting stretch and string stiffness (inharmonicity), and calculates a set of duly compensated frets". For the next few builds at least I have pre-slotted boards, but it might be interesting for those of you looking for better intonation to have a look at it. From what I understand, it doesn't use a constant multiplier to compensate as the effect of string stretch etc. is different at different frets and for different strings.