Make your own free website on

[ Guitars by Cumpiano ] [ Classes by Cumpiano ] [ Newsletter archive ] [ Articles by Cumpiano ] [ Book by Cumpiano ]

William Cumpiano's
String Instrument
Newsletter #8

Go to newsletter number:


Any Questions?

William R. Cumpiano 1998, All Rights Reserved

I've been inordinately busy lately preparing for a flurry of tutorials, so I'll take the opportunity to simply respond to some of the more interesting inquiries I've received recently from some of our newsletter's subscribers.

Melvyn Hiscock wrote:
> Dear Bill,
> My good friend David Kinghorn, a maker of fine acoustic guitars, was asking me what I knew about Kasha bracing. I know very little about it as I am quite happy with other bracing methods and I tend to steer clear of 'perfect solutions' that are 'proved' with unreilable science. The Kasha system may well be fine but I don't want to get involved in the minutae that seems to go with it. I did notice that you were adding words of common sense in the recent spate of newsgroup postings so I thought I would ask you a couple of questions on behalf of internet-less Dave.

The Kasha system was designed for Classical guitars but Dave has seen something that suggests there is a variation for steel strung guitars. Is that so?

> Finally, does it actually work in a way that is noticeable or could the results be down to variations in the top wood, bracing density, etc. in plain English is it that much different to warrant further investigation?

> I look forward to your reply,
> Melvyn Hiscock

I'm not an advocate of the Kasha system, so I'm not the best source of detailed information on it. Michael Kasha has been extolling the system's virtues for over a decade, and through a handful of zealous advocates has been trying to mainstream the system. It just hasn't caught on. He's tried to attain a mantle of scientific respectability for it by submitting the scheme to the Acoustical Society of America but has failed to get that august body of scientists to take his presentations seriously (I've been to conventions of that society where this has happened). Why? Because of the quality of his science. He makes little more than unsubstantiated claims and anecdotal explanations of why his scheme ought to work, but fails to show any data. You can claim anything, but no scientist will take you seriously without data to back you up.

But luthiers require far less stringent proof. He's gotten a lot of enthusiastic press and lot of earnest and proselyte followers, but it just has not caught on. The ear is the final arbiter, and it apparently just does not meet up to that final test. Let alone that it is far, far more complex than standard systems without the obvious increment in sound quality that would justify it.

Now if you're interested in a radical cutting edge system, take a look at the lattice bracing system, as introduced by the Australian luthier Greg Smallman. Now there's a noticeable and dramatic improvement in power with only a modest increment in complexity. I'm experimenting now with transferring the system to steel string tensions. Stay tuned. As for Kasha, I don't know: if you're impressed, follow it to where it will take you. But I am not.

Good luck and best wishes.

Martin Koch wrote:
> Dear William,

> in one of your newsletters you spoke about the proper neck angle. This is a matter which I became aware off very late, despite the description in your book. An eye opening moment was, when I strung up my first dobro with a flat top and a 0 neck angle! You will imagine how HIGH the string action turned out. But I know it since that moment. I had to put off the neck and to build a neck angle into it. But I always disliked the fact that the fingerboard extension has to be pressed down.

>You promised to describe Michael Dresdner's way of building a slope into the upper third of the soundboard. I'm looking forward to this article since this sounds very interesting despite the fact that the building will become more complicated. I don't know how this slope angle is determined since it must be done before the body completement.
> I have some more questions regarding the problem: "what to do with the fingerboard extension on a back angled neck":

> As you know, I like really flat tops. I would also prefer a 90 angle at the neck to body junction but this is impossible with the usual, common bridge height of 3/8". You once said that putting more off the upper fingerboard region when making the radius, sometimes is enough to give the proper neck angle. I haven't tried it yet but could that be a solution for "my" guitars? The neck angle could be 0. The angle is build into the fretboard surface.
> Next "solution": What's about a bridge height that equals the fingerboard height? Why has the bridge to be 3/8" high? Couldn't it be about 1/4" instead? It might happen then that the saddle slot has to go through or almost through the bridge and that the saddle touches the top in the former case, but I don't see a problem with this. I can imagine a 1/4" tall bridge and a saddle that protrudes the optimal 3mm. The sound board is flat, the fingerboard surface straight parallel to the sound board and a ruler on top of the frets runs within the optimal distance (0,4 to 1,6 mm as described in your book) over the bridge.Must a bridge have more mass?
> regards
> -martin

Of course, you can use a 1/4" high bridge, but there are drawbacks:

1- A low string array over the face usually translates to lowered sound output. 2- Since you should have a web of wood under the saddle slot for the bridge's integrity, the slot has to be very shallow (even shallower if you're using a saddle transducer), and the saddle may tip forward. I would only recommend a 1/4" bridge on a small guitar, a 5/16" bridge on a 00 or a 000, and a 3/8" bridge on a Dreadnaught or Jumbo, to drive the soundboards properly. There are exceptions, as always, but my experience bears these dimensions out.

The thicker bridge, of course, means that you have to go to the trouble of tipping the neck back slightly, assuming you want and even-thickness fingerboard and a slender neck. This in turn means you must approximate or calculate the rise above level of the soundboard's upper transversal brace, to meet the rising fingerboard-end. Do you have any drafting experience? If you draft all the parameters at the center of the guitar, you can predict what that rise should be.

A good friend of mine glues the upper transversal brace at the ends, leaving it unglued for several inches at the center, then builds the guitar. When he sets the neck, he will shape an approximate "shim" to slide between the brace and the top until the top "swells" to the appropriate point. When he is satisfied that the fingerboard lays straight and flat, he then rounds the ends of the "shim" so it wont jam, and slides it in with glue (he works through the sound hole, since the transversal is just adjacent to it).

Good luck wrote:
> You mention that very few people know how to set up a classical guitar properly. Can you tell me briefly what is involved?

> Muchas gracias, Tim


It is not complicated; certainly less so than a steel string guitar. Other than set the heights to the minimum at the nut and the desired height at the saddle, there is little more to do: but few people bother to check the octaves to see that they are accurate up the scale. If they do, there is little more left to do than to insure that the tuning machines are operating efficiently (i.e., not binding anywhere, and their backlash as little as possible), that the strings are sliding smoothly through the nut slots and that the frets and fret ends are trimmed and snug. If the octaves are out, then the set up gets more complex.

An electronic tuner will reveal untrue octaves, if your ear is not up to the task of detecting them. But you can pretty well predict if there is an intonation problem by measuring the scale and verifying that the midpoint of the saddle is at the double-the-12th-fret distance, plus no less than 1/16-inch (1.5 mm) or no greater than .10-inch (2.5mm) compensation. The saddle will have to be moved otherwise, which is a repair task.

The classic-guitar G string is often the "rogue" string when it comes to pitch intonation. It tends to behave erratically due to the fact that to vibrate at open G and to maintain a tension similar to the others it has to be more massive and (thus) thicker. Thicker means less flexible. Less flexible means that its "nodes" -- the portions at the ends which don't vibrate because they are too stiff -- is greater than on the other strings. Thus they tend to play sharp, and sharper still as you play up the fingerboard, because the node length remains the same as the string length gets smaller, and the proportion of vibrating to non-vibrating string material between the fret and the saddle is getter greater at every succesive interval. In those sets with a particularly nasty G string, it has to be set up properly: take some 500-grit sandpaper and sand it (yes, sand it) about half a dozen strokes lengthwise right on the guitar (wrap a postage-stamp sized piece of sandpaper around it), between the nut and the twelfth fret and watch the octaves true up before your ears. Just a few strokes is usually enuf.

On some string sets, the G-string is no thicker than it needs to be and it behaves as well, or nearly as well as the others. I haven't made an analysis of the G's in every brand of classical strings, to see which is the most flexible (Savarez reportedly solves the intonation problem by supplying a wound G string, but it introduces a new problem. Such a thin wound string is very fragile) D'Addario has introduced a Zyex (or is it Tynex?) G string (a different polymer from Nylon) which is truly terrific and revolutionary (I should know, Rich Janes and I were the first to recommend the material for use on guitars). Being denser than nylon, it is both harder (wears better) and more massive at a thinner gauge (i.e., more flexible). It makes a wonderfully loud, balanced, clear G string which intonates at the same compensation as the other. Problem solved.



My excellent, and greatly admired friend, Tim White (he's a newsletter member) has generously consented to allowing me to serialize his unprecedented beginner's course in GUITAR ACOUSTICS. This is really a plum for me and for folks who follow my webpage. The course is available in toto in a compendium which he has published of all the back issues of the late, great Journal of Guitar Acoustics. It is a massive tome, which is available from him at If you rather pick it up chapter by chapter in a more relaxed and random pace, read it here. I will be uploading in serialized form it to my webpage. Every time I do so, I will email a chapter to this newsletter list separately.

See you all soon, and it's a privilege to serve.