String tension calculator, with graphic visualization

[Rachmaninoff]

After a couple years, I was doing some tuning changes and I decided to rewrite my string tension calculator.

Features:

- Compares string tension of multiple guitars at once;
- Calculates tension in kg or lbs;
- Plots a line chart comparing the tensions;
- Supports 6, 7 and 8 string guitars;
- Supports multi-scale guitars (fanned frets);
- Provides several pre-configured commercial string packs;
- Allows customization of string packs;
- Provides several pre-configured tunings;
- Allows customization of tunings.

Link:

String Tension Calculator

rodrigocfd.github.io

 

[TheWarAgainstTime]

Great work! I plugged in all the info for a couple of my main guitars/tunings and all of my gauges follow roughly the same shape on the tension graph. Cool to see how changing gauges around to other configurations I've tried in the past may *seem* more correct since they had a more level/balanced reading on the tension graph, but I know they aren't as comfortable to me in practice.

 

[Grindspine]

Very nice! I saved this to favorites for reference. It is a lot more intuitive of a layout than the Stringjoy or D'addario tension calculators I have seen.

 

[ixlramp]

Good idea to visualise the tensions with a graph, especially with a line drawn through the points. I also do this when designing a string set. Yours is the only tension calculator i know of that does this.

It is useful to see the 'tension curve' and i design a custom string set to make this curve 'smoothly varying' with no abrupt changes of direction (traditional sets are often very unsmooth).
I think that the hands can cope better with varying tension if the tension varies as a smooth curve across the strings.

I am wondering what string manufacturer 'unit weight' (mass per unit length) values or tension values are being used. I cannot see them in the code (i might have missed them).

There are two things about your graph i would change (and i apologise for going into detail, i am a long-time tension geek):

1. Your graph does not show 'zero tension', it always vertically stretches the tension range to a fixed vertical size.

This means the graph does not visualise the relative sizes of the tensions, for example: if tension falls from 20lbs to 10lbs from low to high, the fact that the highest string is half the tension of the lowest is not visualised. This would be very useful visual information to get a feel for the string set.

This also means that however 'tension balanced' a string set is designed to be, it always appears equally unbalanced, with a very small imbalance greatly visually amplified.

2. The strings are ordered by string number from left to right.
Traditional string numbering, with the highest string being 'string 1', is very unintuitive and in my opinion a mistake that unfortunately persists. Tunings are almost always thought of and written from low to high: EADGBE.

The order of strings on your graph is the opposite of how tunings are usually thought of and written, and the opposite of how the strings appear on a right-handed instrument either sitting vertical in a stand or being played.

 

[Zulphur]

Great job. Thanks!

 

[ixlramp]

I am wondering what string manufacturer 'unit weight' (mass per unit length) values or tension values are being used. I cannot see them in the code (i might have missed them).

Ignore my comment above. I looked more carefully at the code and i think i now understand how it works.

 

[ixlramp]

As far as i understand it ...

This tension calculator derives unit weight from gauge using a polynomial curve that has been 'curve-fitted' to real-world unit weight data (data perhaps from D'Addario, or perhaps from multiple sources).
There is one polynomial curve used for plain strings, and another used for wound strings.
This allows the calculation of a realistic unit weight value for any possible gauge value, even something in-between and weird like .043 that might not even exist as a manufactured string.

Depending on what real-world unit weight data was used, this tension calculator will be approximate when used for some brands, or approximate for all brands, because unit weight (and therefore tension) for a particular gauge varies slightly between brands.

As a general any-brand any-gauge approximate calculator the design is excellent. This is useful for when you want to approximately calculate tensions for brands that do not publish tension data.

 

[jwade]

This is great! Definitely agree that the graph should be low to high. I'd also suggest that all tunings be included, right now the standard tunings doesn't include F#, G and G#, and the dropped tunings should really include dropped A#, A, G#, G, F#, F, E and D#.

*Edit: for context I'm using the 6 string/28" settings and would benefit from expanded tuning options to be shown*

 

[AkiraSpectrum]

Awesome work!
Going to delve deeper into this soon!

 

[Rachmaninoff]

I am wondering what string manufacturer 'unit weight' (mass per unit length) values or tension values are being used. I cannot see them in the code (i might have missed them).

I pulled the unit weights from D'Addario catalog. I added the link in the project description now.

There are two things about your graph i would change (and i apologise for going into detail, i am a long-time tension geek):

1. Your graph does not show 'zero tension', it always vertically stretches the tension range to a fixed vertical size.

Done.

2. The strings are ordered by string number from left to right.
Traditional string numbering, with the highest string being 'string 1', is very unintuitive and in my opinion a mistake that unfortunately persists. Tunings are almost always thought of and written from low to high: EADGBE.

The order of strings on your graph is the opposite of how tunings are usually thought of and written, and the opposite of how the strings appear on a right-handed instrument either sitting vertical in a stand or being played.

The problem with changing the order is that the calculator allows comparison of guitars with different string numbers. Imagine this complex situation: you have 3 guitars, with 6, 7 and 8 strings. The chart will correctly compare the lightest strings of each. If I change the string order, the comparison would be completely messed up.

Technically it doesn't seem impossible, but it would possibly require a lot of work. I'll study the possibilities later, to see what I can do.

 

[Rachmaninoff]

I'd also suggest that all tunings be included, right now the standard tunings doesn't include F#, G and G#, and the dropped tunings should really include dropped A#, A, G#, G, F#, F, E and D#.

What are the pitches of these tunings?

 

[ixlramp]

The problem with changing the order is that the calculator allows comparison of guitars with different string numbers. Imagine this complex situation: you have 3 guitars, with 6, 7 and 8 strings. The chart will correctly compare the lightest strings of each. If I change the string order, the comparison would be completely messed up.

Just in case there is a misunderstanding, i will clarify ...

I am not suggesting that string numbering order is reversed, i think it should remain as the standard with '1' being the highest pitch string.
Also, i think the order of the strings in the gauge / pitch / tension table is fine because it matches how the fretboard appears to the player and matches fretboard diagrams.

I am only suggesting that the display order on the graph x-axis is reversed so that it appears like this:
6E 5A 4D 3G 2B 1E
But as you write this might be a lot of work, so no pressure from me I appreciate the changes you have already made.

 

[Rachmaninoff]

I am only suggesting that the display order on the graph x-axis is reversed so that it appears like this:
6E 5A 4D 3G 2B 1E

Oh now I get it: reverse the order just in the graph. That's not hard at all.

I implemented the change and it indeed looks more intuitive. Thanks for the suggestion.

 

[ikarus]

wow awesome work. I have bookmarked it.

 

[jwade]

What are the pitches of these tunings?

They would be A#1, A1, G#1, G1, F#1, F1, E1, D#1 and D1

 

[Rachmaninoff]

They would be A#1, A1, G#1, G1, F#1, F1, E1, D#1 and D1

I added the G# and A standard tunings. I'll check the others later, because I need to test the frequencies of all those low notes.

 

[lost_horizon]

Looks great, damn now I can see why my 9-46 B string is so woolly. Probably need a .012 in there...

 

[Tom odd 7]

Nice work, pal.
Congrats

 

[Peter Krasov]

Nice work and perfect interface. But it seems that kg/cm mode does not work correctly. In this mode, the calculator does not seem to convert scale length from inches to cm.

 

[Rachmaninoff]

But it seems that kg/cm mode does not work correctly. In this mode, the calculator does not seem to convert scale length from inches to cm.

It's intentional. Scale lengths are always in inches.