1. Please take 30 seconds to register your free account to remove most ads, post topics, make friends, earn reward points at our store, and more!  
    TalkBass.com has been uniting the low end since 1998.  Join us! :)

Gibson's new patent

Discussion in 'Luthier's Corner' started by pilotjones, Apr 17, 2006.

  1. pilotjones

    pilotjones Supporting Member

    Nov 8, 2001
    In the other patent thread, a few people expressed interest in discussing a new patent that Gibson has been granted. So here goes!

    The patent: http://patft.uspto.gov/netacgi/nph-...2,450.UREF.&OS=ref/4,852,450&RS=REF/4,852,450

    Before I start, let me say a few things:
    - I'm not a patent lawyer.
    - I'm a guy with an engineering background, and an interest in basses, which has led me to some interest in guitar-related patents. (What a nerd!)
    - I'm learning about the patents as I go along. I will always try to point out when something is my opinion/interpretation.

    That being said, there is a huge new patent by Gibson that deals with several issues in building a neck. There are 69 claims in the patent! Best wishes to anyone who fights their way through the whole thing. I've done what I'd call a "detail skim through."

    Here's my summary of some of the major points of the patent. Of course, I might mischaracterize some things in summarizing, but I'll do my best.

    • they give a lot of background material on temperament systems, scales, and prior art.
    • claims PART ONE -- formulas applied to fret placement (note: the "parts" are mine for clarity, and not part of the patent):
    • People from Gibson have done some serious research and calculation regarding strings. They have developed formulas for modelling a real string (as opposed to the standard "ideal string"), with non-zero bending stiffness and longitudinal stiffness parameters, with a defined core size, with compensation for the fact that tension rises as the string is stretched toward the frets, and with compensation for the fact that string linear density decreases as the string is stretched toward the frets. They also account for real (non-ideal) boundary conditions at the ends of the vibrating length of the string.
    • this is the first that I have heard of anyone doing this work.
    • it is my impression that you cannot patent a mathematical formula or equation.
    • there is nowhere in the patent (IMO (in my opinion)) that Gibson is making any claims to the formulas and calculations themselves.
    • they do claim the following: placing frets on a fretboard to the "perfect positions" for a given real string, as determined by using the equations, as opposed to using the standard placement determined by the "rule of 18" or 2^n/12 method, which is what is normally used, based on an ideal string.
    • they include the case where "stretch tuning" is incorporated into the calculations and fret placements.
    • they include the cases where various scale temperaments other than twelve-tone equal tempered (12TET) are incorporated into the calculations and fret placements.
    • they include the cases where the frets may end up nonparallel, curved (if I remember correctly) or even fanned :)eek:) as a result of such fret placement methods.
    • they use the equations to show how "off" the intonation would be for a real string, if the conventional method is used. The even show the result of a standard intonation compensation saddle-move. This is cool information, but is not one of the patent claims. I was impressed at how well the model reflects my experience-- that a properly-intonation-adjusted standard instrument is very close to perfect, up until the 20th fret or so, where it starts to go a few cents sharp.
    • claims PART TWO -- 3D shape of fingerboard
    • they describe, and make claim to, a fingerboard where the shape along any string, and as a result also the 3D shape of the entire surface across many strings, is determined by maintaining a constant angle, at each fret, between (a) the path from that fret to the bridge, and (b) the path from that fret to the next higher-pitched fret.
    • they also make claim to any similar fretboard that (a)involves that same angle, and (b)instead of maintaining the angle at a constant value, uses "a smooth formula" to determine a changing series of angles a you go up the string, thus also producing a particular 2D path under the string, thus producing a 3D surface across the many strings of the fretboard.
    • it is my impression that the PLEK system of fret dressing, used on Ritter and others' basses and guitars, might already use the equal-angles method mentioned two bullets above. They may possibly have a European patent on it. If they do, that patent could possibly apply only to fret dressing as opposed to original manufacture.
    • claims PART THREE -- other areas
    • fretboards combining the "part one" and "part two" items are covered.
    • in what seems to me to be wholly bizarre, they claim (claims # 55 + 56) not only placing frets on a fretboard by method of "part one", and a fretboard created by such method, and an instrument incorporating such fretboard, but also: "producing notes of a musical scale" by making an instrument according to some of the previously claimed methods and "plucking the real string" or "vibrating the real string." :)eek: :eek: :eek:)

  2. knarleybass

    knarleybass Commercial User

    Apr 6, 2005
    Tustin, CA
    Owner of Ulyate Instruments
    This makes my brain hurt
  3. MattMPG

    MattMPG Matt Pulcinella Guitars

    Apr 6, 2006
    Oh Yeah, well I invented Quilted Maple!

    Gibson tried to sue PRS a few years back to get them to stop making single cut guitars! They failed.

    Whatever this crap is, it will fail too.

    If my Gibson hat didn't look so good on me, I'd throw it out.

  4. Scott French

    Scott French Dude

    May 12, 2004
    Grass Valley, CA
    I was under the impression that perfect fret placement for all keys would be impossible.
  5. pilotjones

    pilotjones Supporting Member

    Nov 8, 2001
    Depends what you mean by perfect. Here you are getting into the area of temperament systems. Twelve-tone equal-tempered, versus Pythagorean, just, etc.

    By what I mean by "perfect placement" in this case, is that once one has chosen a temperament system, such as the standard 12TET system, that the fret placement will produce that set of pitches properly.
  6. fookgub


    Jun 5, 2005
    Houston, TX
    I wouldn't be so quick to dismiss this stuff. It looks like Gibson has done some real engineering that could positively affect the intonation and playability of guitars and basses using their ideas.

    I, for one, am all about the advancement of stringed instrument technology. Only two things concern me: 1) will Gibson actually use this stuff on their instruments, or did they just patent it keep others from using it, and 2) is this stuff practical from a manufacturing standpoint (ie: will us players actually ever see it?)
  7. pilotjones

    pilotjones Supporting Member

    Nov 8, 2001
    I'll offer my opinion on the last question: While the patent describes finding the perfect positions for a particular (size-mass-core diameter) real string, one could also figure for an average string that might be used on the instrument. This would result in certain fret positions, for each string. You then have to decide on how to handle the multiple strings. I see three scenarios:

    - To do each string crossing at each fret in the "perfect position", you might end up with non-straight frets. Very tough to manufacture.
    - You could also compromise somewhat, by imposing the condition that all frets be straight. In this case you could come up with a fret layout that is fully or partially fanned, and once you made a jig to cut the pattern, it would be no harder than cutting a f*anned-fret board.
    - You could also compromise even more, by imposing the condition that all frets be straight, and parallel. Compared to a conventional fretboard, you would then be making some slight adjustment to the fret positions, based on the "average position" of the "perfect positions" of all the strings at each particular fret, based on the calculations/equations. Less benefit, but less work. In this case, there is no significant cost involved, since you would just re-space your gangsaws, and cut as usual. This is all that I personally would expect a company like Gibson to actually implement.
  8. pilotjones

    pilotjones Supporting Member

    Nov 8, 2001
    Small notes:
    - This patent was applied for in 1999 and granted in 2000. So it's not as new as I thought.
    - I'm not aware of Gibson doing anything with it.
    - I did a check for Gibson-owned trademarks, and I didn't spot anything that looked to me like it would be used for a guitar using this patent (anything like "Perfect-Fret", "Accu-Fret" etc.)
  9. Geoff St. Germaine

    Geoff St. Germaine Commercial User

    I have seen modelled "real" strings in several physics journals. They generally use waveguide modelling and they can assign whatever boundary conditions that they'd like. If these guys have any applications, it is generally towards digital modelling of guitar sounds rather than having any interest in actually building a guitar.

    When you say that they have developed a formula, is this what it says in the patent? I'm curious if they actually have an analytical equation or if they have actually done some sort of finite element analysis or other sort of modelling.
  10. pilotjones

    pilotjones Supporting Member

    Nov 8, 2001
    It's an analytical solution, not numerical. Here's my summary. Probably not wholly accurate, but not far from.

    They take the equation for a "cable", i.e. the ideal string equation. They combine this with the equation for vibrations in a beam. They look at two boundary conditions, specifically "clamped" and "pinned", which is to say, allowing rotation at the ends, and not allowing rotation at the ends. they then work these two aspects into it, and state that a real string exhibits a behavior of 0.3 and 0.7 (IIRC) between these (that is, taking 0.0 as pinned and 1.0 as clamped). Add water and stir.
  11. pilotjones

    pilotjones Supporting Member

    Nov 8, 2001
    Hi Ham. Nice to hear from you.

    Here's the executive summary of the first group of claims: They've done some mathematical analysis of string behavior (or maybe just read up on it, in light of Geoff's post), and made a patent claim on applying that math to place frets in ideal positions, beyond the standard applying the "rule of 18" or twelfth root of two" methods to place frets.

    Here's the executive summary of the second group of claims: They have made a patent claim on determining the shape formed by the heights of the individual frets, along the length of a string, by mathematical equation.
    No comment! :)
    That's whats going on. Formula being applied to a process to produce an end result.
  12. +1

    but pilot, can you explain to me a little bit about the difference of gibson's fret placement and the (what i consdered) standard rule of 18?
    for example, where would the fifth fret be compared to the fifth fret of the "rule of 18" method?
  13. Long viva unlined fretless, baby!!!!
  14. westland


    Oct 8, 2004
    Hong Kong
    So is this all that different from the Buzz Feiten tuning system which is supposed to also accomodate stretching, etc. of strings to yield a 12-note tempered scale intonation?
  15. pilotjones

    pilotjones Supporting Member

    Nov 8, 2001
    OK, my concept of the BF system: a simple combination of a compensated nut, and stretch tuning.
  16. pilotjones

    pilotjones Supporting Member

    Nov 8, 2001
    The standard rule of eighteen (which is the original older concept, which is more exactly the twelfth root of two method) assumes a 12TET temperament. Using this assumption, it basically says: use a simple formula (formerly division by a number close to eighteen, actually 17.xxxxx; more recently and accurately using factors of 2^n/12) to lay out the frets in what would be perfect positions if the string weren't stretched out of a straight line when fretted, and if the string behaved ideally.

    Since the string really is stretched out of a straight line when fretted and doesn't behave ideally, you must then do a bridge saddle intonation compensation adjustment. This makes then the string come back very close to playing in perfect tune (in that 12TET temperament system) through most of the neck.

    The Gibson method says: develop a complex equation, that includes the effects of string stretching, includes the effects of a real (non-ideal) string, includes any temperament system you like, and includes stretch tuning if you like. Then place the frets in the exact positions governed by this complex equation. There is no intonation compensation necessary because the frets are in ideal places for a real string, instead of the ideal places for an ideal string.
  17. pilotjones

    pilotjones Supporting Member

    Nov 8, 2001
    Not sure I exactly agree with you on this. The way I think that patents work, this is a new method (using this complex formula, that accomodates real-world effects) to lay out frets for a real string, which I guess hadn't been done before, so... valid patent!
    Sounds like it to me too. I'm thinking that either PLEK doesn't hold a US patent, or if they do, it is limited to dressing an existing fretboard, rather than designing a new one. That's conjecture, of course, I haven't researched it.
  18. ok, that makes sense :)
  19. pilotjones

    pilotjones Supporting Member

    Nov 8, 2001
    That fretboard is laid out for just temperament instead of the usual twelve-tone equal-temperament (12TET). But it would still be based on positions for an ideal string, followed by intonation compensation saddle adjustment. What the Gibson patent would do is combine the same set of target pitches (the chosen just temperament system) with equations to lay the frets in the ideal places to produce them, for a real string.

Share This Page