Re: ATM Tolerance in Foucault-testing?

[epabcc@epa.ericsson.se (Bratislav Curcic)]

> ... 'envelope' of tolerance ...

Well, knowing that rho = 1.22*lambda*F/D (where lambda is the
wavelength of light = 550 nm = 0.00055 mm) by substituting it in
Berry's formula (and taking care of moving/stationary source) you'll
see that these formulae are exactly the same. This is also basis of
Milles-Lacroix "tornado" graph. 
The 1/8 vawefront tollerance will have these "X" values halved, while
1/16 (not surprisingly) will be 4 times smaller.

Bratislav