Gibson's new patent

[Frank Martin]

Anyway...

the fact that I had to search for a while and only found a few results, proves that instruments with these fretboards are quite rare.

 

[superbassman2000]

my next question is, are we going to hear or even know the difference?

 

[paintandsk8]

One last thing from me...

It sounds like the new fret positions, as dictated by Gibson's formula method, wouldn't be very far off the positions shown by traditional calculations. How much different isn't known but it couldn't be much and still match a traditional fretboard. So my question now is: Is it even possible to accurately cut a fretslot that is in the "perfect" position and follows the formula? My fretsaw has a kerf that is something like .025". Assuming I accurately cut my slots, I could still have an error of as much as .0125" and still have the slot touching the original position. If the newly discovered fret positions are even that far off the originals, are we really gaining anything here? We can't cut a slot more accurately except by CNC and even that accuracy could be thwarted by an off-center crowning of the fret. And that's another issue - if a fret is hand crowned, could it defeat the new system because of human error? If it can, then this whole thing is just a rhetorical exercise with not much value to the common builder. And my final question, if this new method creates an audible or even detectable difference, will the instrument still be in tune with other instruments that use conventional methods of fret placement?

Well this is of course where the rubber meets the road. It is my personal opinion that part of the beauty in music imperfection. I imagine that even the best fretless players aren't fretting notes anywhere nearly as precisely as the fret placements determined by this equation. This equation is also not perfect itself because it does not account for varying action and string guages etc. A bass with high action is going to need frets with more compensation than one with the action slammed. Not to mention the error introduced in saddle placement, slot cutting, tuning of the strings themselves, fret crowning,... etc. In the end, the amount of error corrected by this method will be completely eclipsed by all the other sources of error in the tuning of any one note. So, in my opinion, yes this is an excercise in theory alone that will never affect us. This patent has been around 6 years and were just now talking about it, buzz feiten compensation has been around longer than that and it still hasn't caught on widely.

On the other hand I do find this information extremely interesting, and I do think that a good understanding of differences between what's actually "perfect" and what is accepted is good knowledge for any luthier.

 

[Suburban]

Not having gone into depth with the patent itself, I do have a few comments

The "real string" formulas will need a set of assumptions. the most obvious, as it deals with the stretched string, is that the action will impair the intonation due to different stretch.
And as we know, factors like humidity and temperature will impact the action. Hence, all this good work will end uo useless, or at least redundant.

What is worse is that they have claims on the fretboard shape.
You may argue that the patent covers only a method, but any infringment charges will be based on the "thing", since there is no way to show that a method is used, unless it includes dedicated and specialized machines.

Hence, the new approach of the USPAT, to meet the patent fury of Asian countries (to grant the Great Companies more or less anything that shows differences from what this specific patent engineer/lawyer knows of) will end with stolen ideas and crippled small scale inventors in the US. As it looks right now, most European patent offices are moving in the same direction. We shall soon all be working for Big Brother, giving our brain and hand capacity to the Great Companies for free - or be renegades.
Which is for you, friend?

 

[pilotjones]

...
If the newly discovered fret positions are even that far off the originals, are we really gaining anything here? We can't cut a slot more accurately except by CNC and even that accuracy could be thwarted by an off-center crowning of the fret. And that's another issue - if a fret is hand crowned, could it defeat the new system because of human error? If it can, then this whole thing is just a rhetorical exercise with not much value to the common builder.
...

Great, and pertinent questions. Numbers later in the day.

 

[Showdown]

My question would pertain more to guitar than bass, although it would still apply to bass. How would curved frets affect string bending? It seems like if there was much of a curve, and if each fret had a slightly different curve, it would change how far you would have to bend to reach say a half step depending on what fret and even what string you were bending. That would make bending up to exact pitch very difficult because each note would potentially take a different amount of bend to go up the same amount. And what if the curve in the direction you were bending went toward the nut, thus lowering the pitch at the same you were raising it, somewhat cancelling the bend?

As someone else said, this makes my head hurt.

 

[pilotjones]

OK, here goes:

For a 34" scale instrument. Effect of having the fret, or the crown of the fret, too close to the bridge, by a distance of .0125" [.3175mm].

1st fret 0.7 cents sharp
12th fret 1.3 cents sharp
24th fret 2.5 cents sharp


From what I understand, 2.5 cents is getting just into the area of a noticeable difference.

 

[pilotjones]

My question would pertain more to guitar than bass, although it would still apply to bass. How would curved frets affect string bending? It seems like if there was much of a curve, and if each fret had a slightly different curve, it would change how far you would have to bend to reach say a half step depending on what fret and even what string you were bending. That would make bending up to exact pitch very difficult because each note would potentially take a different amount of bend to go up the same amount. And what if the curve in the direction you were bending went toward the nut, thus lowering the pitch at the same you were raising it, somewhat cancelling the bend?

As someone else said, this makes my head hurt.

Sounds right to me.

This is related to bending on a multiscale board. F*nned-fret players learn and get used to the fact that depending where they are on the neck, bending one way or the other will have increased or decreased effect. In fact, it may be possible on the upper frets to actually bend flat instead of sharp by pulling towards the treble side. That would depend on the pitch going flat due to pulling across the fret angle and lengthening the effective speaking string, faster than it goes sharp due to stretching the string. Maybe a Dingwall owner could tell us whether this actually works.

 

[pilotjones]

As paintandsk8 pointed out, applying the formulas to find "perfect" fret positions depends on assuming a particular action (neck shape and bridge height), and a particular set of strings. The fret positions will not be dead perfect if these change. But they could possibly be better than if you didn't take account of these factors at all.

This discussion has me thinking: You could make a bass with frets on the large side, set it up with and average action and strings, and intonate it. You could then do measurements with a precision tuner and map out the errors across the fretboard. It's then pretty simple to figure the exact distance the fret should move to land exactly in pitch. You could then re-crown the frets, moving the crowns backwards or forwards (off-center) as necessary to get perfect pitches. You'd lose some height in re-crowning, but a bridge lowering would mostly take care of that.

I wonder if there are any people who have already done this.

 

[brianrost]

Gibson has done some real engineering that could positively affect the intonation and playability of guitars and basses

Well, it's about time considering their prices

 

[paintandsk8]

OK, here goes:

For a 34" scale instrument. Effect of having the fret, or the crown of the fret, too close to the bridge, by a distance of .0125" [.3175mm].

1st fret 0.7 cents sharp
12th fret 1.3 cents sharp
24th fret 2.5 cents sharp


From what I understand, 2.5 cents is getting just into the area of a noticeable difference.

Wait, is the the amount of error in a regular board as claimed by the patent, or just an arbitrary example of what kind of intonation error can be caused by improper frets?

 

[Sheldon D.]

Sounds right to me.

This is related to bending on a multiscale board. F*nned-fret players learn and get used to the fact that depending where they are on the neck, bending one way or the other will have increased or decreased effect. In fact, it may be possible on the upper frets to actually bend flat instead of sharp by pulling towards the treble side. That would depend on the pitch going flat due to pulling across the fret angle and lengthening the effective speaking string, faster than it goes sharp due to stretching the string. Maybe a Dingwall owner could tell us whether this actually works.

I've tried this and no matter what string or fret angle, the pitch always goes up. IMO the angle of the fret would have to be very steep for the increasing length to overtake the increasing tension and cause the pitch to drop.

 

[Geoff St. Germaine]

As Sheldon has said, the pitch always goes up.

Peter, you should be able to calculate the angle at which bending would work the other way.

 

[Marcus Willett]

OK, here goes:

For a 34" scale instrument. Effect of having the fret, or the crown of the fret, too close to the bridge, by a distance of .0125" [.3175mm].

1st fret 0.7 cents sharp
12th fret 1.3 cents sharp
24th fret 2.5 cents sharp


From what I understand, 2.5 cents is getting just into the area of a noticeable difference.

Very interesting though (as Ham suggested), probably academic stuff. What's most interesting to me is that this whole area of discussion involves making compensations/corrections that might at most amount to a couple cents. Uh huh. Problem is that even if the instrument is perfectly in tune (assuming 12TET) on every note on the instrument, you're still out of tune (relative to the "ideal"), and many notes much moreso than a couple cents. I mean, what's an even-tempered major 3rd off by...? It's like 15 cents (IIRC), or something in that area.

(Kinda reminds me of a discussion I saw years ago about getting sample-accurate sync and how critical it was. then I heard an old-school engineer say something along the lines of "It's funny...for years we synced 24 track tape machines together with far less accuracy and it never seemed to be a problem.")


12TET is a compromise in and of itself. You can't perfect that...ever.

In the end, the amount of error corrected by this method will be completely eclipsed by all the other sources of error in the tuning of any one note.

I think this is a very good summary.

It is interesting tho....

 

[pilotjones]

As Sheldon has said, the pitch always goes up.

Peter, you should be able to calculate the angle at which bending would work the other way.

Yes, I could. Too lazy. Plus experience beats analysis.

 

[pilotjones]

Wait, is the the amount of error in a regular board as claimed by the patent, or just an arbitrary example of what kind of intonation error can be caused by improper frets?

Arbitrary example, based on Hambone's discussion of "what if the kerfs were off by one-half a kerf width."

 

[pilotjones]

I've tried this and no matter what string or fret angle, the pitch always goes up. IMO the angle of the fret would have to be very steep for the increasing length to overtake the increasing tension and cause the pitch to drop.

Want to see very steep, and able to cause a bend up or down? Look at patent #5,133,239:
Invalid Link Removed
You've got to see the images.

I doubt anyone has actually done this. If the nut slot or the bridge saddle were the tiniest distance off to either side, the string's pitches would be entirely off.

 

[pilotjones]

Problem is that even if the instrument is perfectly in tune (assuming 12TET) on every note on the instrument, you're still out of tune (relative to the "ideal"), and many notes much moreso than a couple cents. I mean, what's an even-tempered major 3rd off by...? It's like 15 cents (IIRC), or something in that area.

12TET is a compromise in and of itself. You can't perfect that...ever.

True. But so is just intonation, even in its home key, with the Pythagorean comma and all.

The only way I know to have no inharmonicity is to continuously adjust the overtone series pitches of each note, depending on the current context. William Sethares does it using computers. Website: http://eceserv0.ece.wisc.edu/~sethares/

 

[Marcus Willett]

True. But so is just intonation, even in its home key, with the Pythagorean comma and all.

Exactly, which is more-or-less my point. I think it's important as a musician to never lose sight of the basic axiom that perfection can never be achieved; it is the attempt to achieve it in and of itself that yields the rewards.

 

[Tim__x]

OK, here goes:

For a 34" scale instrument. Effect of having the fret, or the crown of the fret, too close to the bridge, by a distance of .0125" [.3175mm].

1st fret 0.7 cents sharp
12th fret 1.3 cents sharp
24th fret 2.5 cents sharp


From what I understand, 2.5 cents is getting just into the area of a noticeable difference.

For a sequence of non simultaneous notes, three cents is about the limen for an average person. The trained ear is a little better than that.

For simultaneous intervals, much much less is detectable, under the proper conditions even a fraction of cent is perceptible. Of course, barely perceptibly different doesn't mean it's worse or unpleasant.