Optimal string tension

[Sollipsist]

Has anyone done a serious study of tension, string gauge and scale length? I know just enough of the physics to understand that sustain increases with tension to a point, and then decreases after that. So there must be a fairly small range of variables at which any given string gauge tuned to a given pitch is reaching its potential.

Obviously the different scale lengths are designed to offer a close compromise. But has anyone done the work to see how close they really are?

Of course this is leaving aside all of the other considerations -- fretting, composition of the strings, how they're anchored on either end, magnetic string pull, the quagmire issue of tonewoods, etc... I'm just looking for as simple a test as possible on the best achievable tension, gauge and length formula.

 

[MaxOfMetal]

The issue will always come down to the player.

The "best" sounding strings and scale might not be the most comfortable or conducive to said player's technique.

Strings are so damn cheap. Try every set you can get your hands on, even if you're not sure you'll like them. That's part of the fun.

 

[Sollipsist]

The issue will always come down to the player.

The "best" sounding strings and scale might not be the most comfortable or conducive to said player's technique.

Strings are so damn cheap. Try every set you can get your hands on, even if you're not sure you'll like them. That's part of the fun.

Absolutely, after a LOT of trial and error, I finally got a "perfect" custom Stringjoy set for each guitar. I'm now looking at it from the other end -- how to build or find a guitar that's closer to ideal already, hopefully cutting down the trial and error time at least a little.

 

[MaxOfMetal]

Absolutely, after a LOT of trial and error, I finally got a "perfect" custom Stringjoy set for each guitar. I'm now looking at it from the other end -- how to build or find a guitar that's closer to ideal already, hopefully cutting down the trial and error time at least a little.

The problem is, you can't exclude variables that are inconvenient.

 

[Sollipsist]

The problem is, you can't exclude variables that are inconvenient.

Fair enough, but it's acceptable to exclude variables when testing for the effect of specific other variables

Obviously no given gauge string is going to perform well unless it's within a fairly narrow range of pitch and the necessary tension to achieve that pitch. So it stands to reason that there's an ideal point of tension for that string, and any variables other than gauge and scale length wouldn't necessarily change that. I'm just wondering if anyone has established such a reference, and how close our designs are to that perfect reference point. Or am I looking at the whole thing the wrong way?

 

[MaxOfMetal]

Fair enough, but it's acceptable to exclude variables when testing for the effect of specific other variables

Obviously no given gauge string is going to perform well unless it's within a fairly narrow range of pitch and the necessary tension to achieve that pitch. So it stands to reason that there's an ideal point of tension for that string, and any variables other than gauge and scale length wouldn't necessarily change that. I'm just wondering if anyone has established such a reference, and how close our designs are to that perfect reference point. Or am I looking at the whole thing the wrong way?

Material and construction differences between like gauge strings can be significant.

From different chromium content in the steels to the purity of the nickel, from hex core to round core, how large or small the windings are, etc.

I feel you are sort of looking at this the wrong way.

 

[Sollipsist]

Material and construction differences between like gauge strings can be significant.

From different chromium content in the steels to the purity of the nickel, from hex core to round core, how large or small the windings are, etc.

I feel you are sort of looking at this the wrong way.

I see the point. If we only used one kind of string, it would be much more easily tested.

But on the other hand, we're using all of these different strings on basically only three or four scale lengths, and there's a rough consensus that some combinations work better than others for a majority of players. It may be unachievable, but it's tempting to think that there's a scientifically testable ideal range that can be determined.

From all I've heard about Leo Fender, it seems possible that our standard 25.5" scale is based on nothing more than it was the most cheap and easily available option at the time

 

[MaxOfMetal]

I see the point. If we only used one kind of string, it would be much more easily tested.

But on the other hand, we're using all of these different strings on basically only three or four scale lengths, and there's a rough consensus that some combinations work better than others for a majority of players. It may be unachievable, but it's tempting to think that there's a scientifically testable ideal range that can be determined.

From all I've heard about Leo Fender, it seems possible that our standard 25.5" scale is based on nothing more than it was the most cheap and easily available option at the time

The scales we use are based on little more than tradition and pop culture. There wasn't a scientist in 1160 A.D. experimenting with scales when they made one of the earliest instruments that would eventually become the guitar. They just based it on their gut and whatever material and construction constraints existed at the time. Not to mention they definitely weren't measuring in inches.

 

[Winspear]

It's really incredibly subjective.
Inharmonicity is one thing that can be measured and is fairly agreeable across the board. That's the effect of thick strings on short scales, that cannot vibrate properly, feel dead, have dull tone and out of tune overtones. I think most people are in agreement that 7th strings around mid 60s start to sound bad on standard scale. Likewise people disliking gauges 80 and up on 8 strings and such. (What they can tune that string to with acceptable tension is up to them - the tone remains the same).
More scale length for a thinner gauge and more brightness is the obvious solution here, but there's not really much math to it. Sure, the inharmonicity can be calculated but what kind of performance is acceptable for the string is a matter of preference.
That (inharmonicity/clarity) is just one element, without taking into account tension, fret buzz, action etc and all peoples different preferences.
Don't overthink it.

Find your minimum string tension. Expect it to be slightly higher on longer scales, because longer scales give the string more play to feel flexible (the reason they sound clearer).
Pick your desired tuning. Use a gauge that gets your minimum string tension. If it's too light and buzzy, either raise the action or use a heavier gauge. If its too dark, you need a longer scale for your preference.
All there is to it really!

 

[Politics of Ecstasy]

Tom, I dont mean to pick your brain, but how does inharmonicty happen? Or is that like intonation issue?

Meaning, lets say I tune my 7620 lo pro edge to B standard.....now when I play in the. Middle of the fretboard, some notes seem to be “wrong” even though I’m playing the “right” note

Is that intonation or inharmonicity?

 

[Winspear]

Inharmonicity happens when the string can't vibrate properly because it's too big/stiff for its short length. It will lead to bad intonation (the word here meaning tuning/accurate note production - this can happen even on a guitar that is intonated as perfectly as possible) - it's why the thickest strings sound bad and aren't in tune on the upper frets. It's also why thick plain Gs sound bad for the same reason.

If some notes are off in the middle of the fretboard, it could just be bad intonation (the word this time in the typical guitarist sense - wrong pitch due to setup issues, saddle position, action, relief..).
In my experience there's no reason for a 25.5 to have poor intonation with a gauge up to mid 60s. If the string is tight enough, action low enough, saddle positioned properly, then I am not sure why those frets are out.
It's not inharmonicity. That'll by why the 60+s start to sound dark, but they don't really suffer from poor tuning until right up the board.
Unless you are calling middle of the fretboard like 14+ where that string will start to go out , in my experience

 

[c7spheres]

I'm curious. Is anyone actually using a tension guage when doind setups or tuning thier guitars? If so, what are you using and is it worth it? I see them for sale but never met anyone who owned one (not even techs I know).

 

[ixlramp]

As far as i know, the tension equation that uses 'unit weights' (that all the tension charts and calculators use) gives results that are fairly close to reality, such that it is not worth actually scientifically measuring tension.

 

[KR250]

I'm curious. Is anyone actually using a tension guage when doind setups or tuning thier guitars? If so, what are you using and is it worth it? I see them for sale but never met anyone who owned one (not even techs I know).

I use a string calculator, typically the same one so I have consistency when ordering strings. Once I string up a guitar and play it I'll make notes on what I feel is too much/little tension or too dark sounding then update my preference when ordering the next set. Trial and error I suppose. More times than not, when starting a custom guitar build I begin with the tension calculator and desired tuning and build around that (especially with multi-scales).

 

[bostjan]

Inharmonicity happens for a number of reasons: because the string gets so thick that it can no longer be modeled realistically as a "perfect string" of infinitescimal length, and starts to vibrate like a rod fixed at both ends, because the fixed ends of strings start to yield, because strings are not uniform along their length, but most of all due to poor elasticity per unit length (especially wound strings). All four factors contribute to an 8 string guitar with a trem.

For thin strings, they tend to sound optimal just before they start to yield, which is actually not dependent on thickness, so gauge doesn't really get determined by that property.

For thick strings, inharmonicity starts to be a factor, leading to thinner strings being a better choice. But if a string has less tension than the player is comfortable fretting, it won't sound good. So, in that sense, it's subjective. Plus, some players enjoy making the "bwow" sound you get from picking hard enough to bend the attack sharp, and others enjoy the inharmonic metallic thwack of overly thick and tight strings.

Forgetting about practicality for a minute, the clearest tone comes from having a scale length long enough to bring your highest string closest to yielding, and then using tension on the lowest string just above the point where pick attack and fretting would cause noticeable deviation in tone (so use a light touch). For a high E, that's around 30" scale, but that sort of scale, most players agree, is too long for comfort, and 99% of guitarists like to bend their strings, so it has to be shorter than that in practice.

 

[Politics of Ecstasy]

Inharmonicity happens for a number of reasons: because the string gets so thick that it can no longer be modeled realistically as a "perfect string" of infinitescimal length, and starts to vibrate like a rod fixed at both ends, because the fixed ends of strings start to yield, because strings are not uniform along their length, but most of all due to poor elasticity per unit length (especially wound strings). All four factors contribute to an 8 string guitar with a trem.

For thin strings, they tend to sound optimal just before they start to yield, which is actually not dependent on thickness, so gauge doesn't really get determined by that property.

For thick strings, inharmonicity starts to be a factor, leading to thinner strings being a better choice. But if a string has less tension than the player is comfortable fretting, it won't sound good. So, in that sense, it's subjective. Plus, some players enjoy making the "bwow" sound you get from picking hard enough to bend the attack sharp, and others enjoy the inharmonic metallic thwack of overly thick and tight strings.

Forgetting about practicality for a minute, the clearest tone comes from having a scale length long enough to bring your highest string closest to yielding, and then using tension on the lowest string just above the point where pick attack and fretting would cause noticeable deviation in tone (so use a light touch). For a high E, that's around 30" scale, but that sort of scale, most players agree, is too long for comfort, and 99% of guitarists like to bend their strings, so it has to be shorter than that in practice.

Great reply and info
Thank you man for taking the time to post

 

[ixlramp]

Sollipsist, in the context of ignoring many factors and focussing on particular optimisations, it depends on what you are trying to optimise, is it sustain only, or sustain and harmonicity, or what?

For optimising harmonicity, longest possible scales for minimising gauges and therefore stiffness:
Longest practical scale for top E4, which works out at around 30".
Then, as extreme a fan as is practical and playable to extend the lower strings longer than 30".
Also worth mentioning: Avoid large gauge plain steel strings (anything over .016, even .016 seems a little stiff to me), try to use the thinnest practical roundwound at the transition from roundwound to plain steel.

For sustain i suspect that the acceptable range of tensions is fairly wide, not narrow.

But when it comes to an actual guitar design, the number of factors involved and the balancing of these makes me suspect that any isolated experiment and results would be almost useless to know. However i do not want to discourage investigating isolated optimisations as they are interesting and will always have some kind of benefit.

 

[Sollipsist]

Sollipsist, in the context of ignoring many factors and focussing on particular optimisations, it depends on what you are trying to optimise, is it sustain only, or sustain and harmonicity, or what?

For optimising harmonicity, longest possible scales for minimising gauges and therefore stiffness:
Longest practical scale for top E4, which works out at around 30".
Then, as extreme a fan as is practical and playable to extend the lower strings longer than 30".
Also worth mentioning: Avoid large gauge plain steel strings (anything over .016, even .016 seems a little stiff to me), try to use the thinnest practical roundwound at the transition from roundwound to plain steel.

For sustain i suspect that the acceptable range of tensions is fairly wide, not narrow.

But when it comes to an actual guitar design, the number of factors involved and the balancing of these makes me suspect that any isolated experiment and results would be almost useless to know. However i do not want to discourage investigating isolated optimisations as they are interesting and will always have some kind of benefit.

A lot to work with here, thanks so much.

I'm trying to wrap my head around the idea that a longer scale improves the harmonicity of higher strings more "quickly" -- in other words, a 27" scale is closer to ideal for the 1st string than it is for the lowest string. Which would imply that, if harmonicity was your chief concern, you'd be better off with a straight 27" than with a 25.5"-27" fan?

 

[ixlramp]

a longer scale improves the harmonicity of higher strings more "quickly"

Yes, string stiffness is determined by its gauge relative to its length. A .009 on 27" is very thin relative to its length, stiffness is not an issue, whereas a .080+ on 27" has significant stiffness problems.

if harmonicity was your chief concern, you'd be better off with a straight 27" than with a 25.5"-27" fan?

Yes, although ... the shortening mostly occurs for the higher strings, where string stiffness is not much of a problem, so it is not a huge advantage.

One advantage of making the higher strings longer is that the thickest plain steel can be made thinner, reducing its stiffness problem. For example .015 instead of .016.

 

[bostjan]

Especially with extended range, the inharmonicity is going to be so much larger on the thickest bass string that the others could be neglected.