Official Luthiers Forum!

Yeah, sorry about that folks - it's been a long and busy week, and I spoke before thinking there. The numbers and formula I spit out were related to calculating impact of fret placement error, not adjusting saddle compensation. For saddle compensation i do actually use an equation pretty much as Frank laid out, just arranged a bit differently. So double that number to 14.

That's the problem with using spreadsheets - I work out the formulas once, then spend years getting lazy and forgetting them as the computer does the work for me. D'oh!

_________________
Eschew obfuscation, espouse elucidation.

Is there a similar calculation for nut compensation?

Yes, there is. But you have to use it simultaneously with a saddle compensation equation and decide on which fret or frets you want to reduce the intonation error to zero. Compensation at the nut affects all frets an equal number of cents whilst compensation at the saddle affects all frets differently (the same compensation amount is applied to different string speaking lengths). The complication is not so much in figuring out the tonal shifts but modeling the neck relief geometry in a way that relates to how action and relief are normally measured, because to solve the compensation equations properly you need the action for each string at every fret of interest and then minimise the sum (across all strings, all frets) of all the intonation errors.

_________________
Trevor Gore, Luthier. Australian hand made acoustic guitars, classical guitars; custom guitar design and build; guitar design instruction.

http://www.goreguitars.com.au