Compensated saddles?

Subject: Compensated saddles?
From: Scorpio
Date: 04 Jul 10 - 03:58 PM

Why do manufacturers use compensated instead of straight saddles? Does the advantage depend on the guitar? Scale length? Personal taste? Is Compensated always better?

Subject: RE: Compensated saddles?
From: GUEST,DonMeixner
Date: 04 Jul 10 - 04:52 PM

Most classical guitars have straight saddles and most steel stringers are compensated. There many reason but they all have to due with string diameters and string intonation. Hang on and the scientists will arrive and straighten it all out.

Don

Subject: RE: Compensated saddles?
From: open mike
Date: 04 Jul 10 - 05:03 PM

so this is not about horseback riding, then?

Subject: RE: Compensated saddles?
From: Brian May
Date: 04 Jul 10 - 05:04 PM

Try here:

http://www.frets.com/fretspages/general/glossary/Comp/comp.html

This is a luthiers' website (guitar makers etc).

Subject: RE: Compensated saddles?
From: Richard Bridge
Date: 04 Jul 10 - 05:38 PM

It is simple science. Owing to the stiffness of the string the sine wave behaves as if the node is not actually at the saddle (or nut - the thinking behind Buzz Feiten compensation) but at a distance in front of the saddle (or nut) that is determined by the relationship between mass and stiffness. The concept is simple but the detailed maths would be hideous so trial and error is the best way. Nylon strings are much less stiff so the problem is not so marked.

Subject: RE: Compensated saddles?
From: GUEST
Date: 04 Jul 10 - 06:23 PM

What Richard said. But said more simply,when guitarists playing steel strung guitars 'bend' notes they push the string sideways, stretching it, to raise the pitch of the note. Just pressing the string down onto the fret does the same thing but to a lesser extent. To counteract this the bridge is compensated, i.e. the point where the string touches it is moved a little further away from the nut relative to its theoretical position. The stiffer the string, the more compensation is needed. Classic guitars have relatively floppy strings needing little or no compensation. Electric basses have very thick strings with very stiff steel cores so they need a lot.

Subject: RE: Compensated saddles?
From: Murray MacLeod
Date: 04 Jul 10 - 07:21 PM

An interesting article here on the compensated nut. Well worth a read.

Most saddles on steel strung guitars are slanted, which is an attempt, albeit half-hearted, to address the compensation issue.

Some steel strung guitars, (Lowdens for example) have a two part saddle, which is an improvement, but the only way to get as near perfect intonation as possible is by painstakingly shaping the saddle for each individual string, with the aid of a Peterson strobe tuner (any other type of tuner is useless for the purpose)

It would also be beneficial to obtaining correct intonation if saddles were thicker than most manufacturers supply. 3.5mm is a minimum thickness as far as I am concerned. The problem here of course is that a thick saddle is less attractive visually ...

Assuming the frets have been correctly located, it is possible to get the guitar to play as near perfectly in tune as humanly possible by compensating both the nut and the saddle, but it is a long slow process.

Worth it in the end, however, IMHO.

Subject: RE: Compensated saddles?
From: JohnInKansas
Date: 04 Jul 10 - 07:55 PM

Electric basses have very thick strings with very stiff steel cores so they need a lot.

The compensation needed is largely due to the "wound string" construction of the strings.

"Stiffness" is commonly used to indicate the resistance to bending and all strings are intended to be as "limp" as possible with respect to bending. Heavy (large diameter) strings would be "stiff" in bending, so a thin wire is used in the center and a heavier wire is wound around them to increase their weight.

With the amplification available with an electric guitar, usually a smaller core wire can be used to get lower string tension, with a fatter winding to get the mass per unit length. The lower tension facilitates "lots of bends" and string amplitudes actually can be lower so a "fatter winding" can be used. The compensation needed is a matter of geometry rather than "stiffness."

The "pitch" of a vibrating string depends (theoretically) only on the length, tension and mass (loosely, weight per unit of length).

Most string materials, but particularly steel, have a rather non-linear relationship between "stretch" and tension at low tension (low stress in the tension-bearing part of the string) so they produce a "flabby" sound that damps out rather rapidly at low tension. To get a "bright" sound, the core of the string that carries the tension needs to be (for steel) at, or slightly above the nominal value of 80% of the breaking strength of the string.

To get a solid string with sufficient mass to produce the low tones wanted on some of the strings, 80% of the breaking strength would exceed the breaking strength of the instrument, so a "weak" core wire has its mass per unit length increased by the wire wound around it.

As noted above, the "node" (point of zero motion) at the bridge isn't necessarily at the point of contact, and the displacement when the string is pulled sideways to fret (or bend) moves more for a large diameter (wound) string than for a skinny one, so the "effective string length," for a straight bridge, is slightly different for "fat strings" than for "skinny" ones. The displacement of the node point is more a matter of geometry than of string tension or "weight."

The compensated bridge adjusts for the difference.

Theoretically if one really wanted a multi-string instrument to play everything exactly in tune the proper solution would be to put each of the frets in the right place for each string, with different fret spacings for each string - since the strings "want to be" of different length if bare and wound strings are mixed together; but the (theoretical) change in fret location isn't linear, or even a continuous curve across the width of the fingerboard with string combinations commonly used. People have tried this method of "compensation" but it makes a very ugly instrument, and doesn't really do much better than just adjusting the physical string length via a compensated bridge to make the effective string lengths all "nearly" the same.

An alternative theoretical approach would be a set of strings that all have identical core-wire diameter, with each string wound with a sufficient density winding to get the appropriate open pitch for that string. This also "didn't really work" for those who report trying it. The fret locations are (theoretically) better, but compensation for the huge variation in string diameters is even more necessary.

The compensated bridge usually is more necessary for short string instruments than for those with longer strings (a matter of "percentage deviation") but even for a long-string instrument benefits may be seen for those who play "close to the bridge."

If you don't play much "past the middle" you probably are unlikely to improve how your guitar sounds when you play it by adding compensation to the bridge; but as with all things "instrumental" it's a method that's available if you can find a way to improve your satisfaction with your instrument when you play it. (With due consideration to how it sounds when you loan it to someone with different technique, if you like the borrower and it happens often enough?)

John

Subject: RE: Compensated saddles?
From: Scorpio
Date: 04 Jul 10 - 08:18 PM

More info than I can shake a stick at. Thanks scientists.

Subject: RE: Compensated saddles?
From: Richard Bridge
Date: 05 Jul 10 - 04:20 AM

I partly agree with what John says, but the bend-related intonation problems go away almost entirely if the action is low enough -I play 1.9 (treble) 2.7 (bass) (mm) and I'd play lower if I didn't seem to have an incurable addition to really digging in on the bass side.

As fat as I can see intoning the nut is a waste of time (subject as follows) since all of the strings will be fretted some of the time. Nut intonation merely (usually, IMHO) masks the effect of the nut being too high, very often on the G string.

Intoning each fret is not practical. But I do have a 12-string that does play pretty nearly in tune almost all the way up the neck (and each course stays in with itself almost all the way up the neck) because, as Murray says, each individual string was separately intoned using a Petersen (by Brian Rodgers to whom many thanks).

Capos of course cause tuning problems, and the best resolution is to have a capo that matches the camber of the fretboard (which is difficult since most fretboards these days have compound camber ie the fretboards get flatter the higher you go, and also because dim luthiers round the ends of the frets off too gradually so that the top and bottom E strings demand the capo be overtightened: rounding should be outside the area of play) and then to apply the Fielding rule - ie capo as close as possible to the fret (even on the back half of it) and as loosely as possible so long as the string is held to the fret and the capo does not fall off - to minimise string bending..

Subject: RE: Compensated saddles?
From: Murray MacLeod
Date: 05 Jul 10 - 05:06 AM

A minor point, Richard, but it is by no means the case that most guitars these days have compound radius fretboards.

Among factory built guitars, ie Martin, Taylor and Gibson, the fretboards are still universally produced with a constant radius, 16" in the case of Martins, 15" in the case of Taylors.

Making a compound radiused fretboard (or "conical" radiused to be strictly correct terminology -wise) is a much more time consuming and labour intensive operation than making a constant radiused fretboard, ergo ...

Subject: RE: Compensated saddles?
From: JohnInKansas
Date: 05 Jul 10 - 05:55 AM

With the kinds of equipment that most luthiers who are doing it as a business are likely to have (or to have built) there should be virtually no difference in the amount of time or labor rquired to make a constant radius (cylindrical) fretboard or a tapered (conical) one. Both shapes are what geometers would call "generated surfaces," and an exceedingly simple machine (or call it a jig or a fixture) setup can easily, and accurately, create either kind.

If, for some reason I'd have difficulty understanding, you'd like an elliptically tapered fingerboard, or one that's exponentially tapered, that's almost as easy as one that "looks simpler."

The cross-curvature of the fingerboard, however, has relatively little to do with the need for a compensated bridge, unless perhaps a particular curve encourages (or discourages) more or less use of large "bends" in your playing. That's a "how you play it" sort of thing, and as suggested above I'll again assert that you should do whatever gives you a warm feeling when you play your instrument (within the social tolerance of your fellow musicians and audiences).

The need for a compensated bridge is primarily due to the inability, with readily available strings and structures, to have all the strings the same diameter so that they all have the same "end effect" at the bridge. As an alternative to having all the strings the same diameter you might have the diameters increase linearly across the bridge so that a slightly "slantwise" straight bridge could compensate "correctly." In that case the "slant" likely could be slight enough to be almost invisible without accurate measurements; but unfortunately no such set of strings is very practical for other (probably more important) reasons.

John

Subject: RE: Compensated saddles?
From: Murray MacLeod
Date: 05 Jul 10 - 06:09 AM

John, I am talking about factories turning out hundreds of guitars a day, not individual luthiers producing one or two a month.

A constant radius fretboard in a factory set-up like Martin or Taylor is produced in a matter of seconds by feeding it through a shaper (or spindle moulder as we Brits call it) with the appropriately ground cutter in the arbor.

You can't make a compound radius fretboard as fast as that.

Hence the reason why factories prefer constant radius fretboards.

(And also, of course, because it makes the installation of the frets so much simpler for semi-skilled operatives).

Subject: RE: Compensated saddles?
From: Richard Bridge
Date: 05 Jul 10 - 09:10 AM

Well, Murray, if that is so it points out another problem.

The strings are always wider apart at the saddle than at the nut so in the absence of a compound radius fingerboard then either the fret heights will change to create a compound radius notional playing surface, or the playing clearance at the frets will not progress correctly up the neck - I think it will be the bottom and top E will find (dis)proportionately greater fret clearances as you progress up the fingerboard.

I almost can't believe that to be the situation.

Subject: RE: Compensated saddles?
From: Murray MacLeod
Date: 05 Jul 10 - 09:51 AM

It is indeed the situation, Richard, and I find it difficult to understand how the major companies can be so reactionary as to persist with constant radius fingerboards, when the benefits of the compound radius are so obvious. Well actually I don't find it difficult to understand, I know what the accountant mentality is like ...

It is less of a problem on acoustic guitars, and the flatter the radius, the less of a problem it becomes, but the original Fenders, Teles and Strats, were a nightmare, with a constant 7 1/2" radius it was impossible to bend a note cleanly above the 12th fret.

Hence why Warmoth took the replacement Fender neck market by storm when they introduced their compound radius replacement necks.