Re: ATM Foucault test accuracy

["Tom & Lou Krajci" <krajcit@3lefties.com>]

> From: "Bill T." <twentiethwave@hotmail.com>
>
> If a barely gray shadow is a "1" and an almost black shadow is a "10", I
get
> my best *repeatability* with a bright shadow of about 3.  My worst
> repeatability is for shadows darker than about 6....

This reply, and others like it, keep bringing up one aspect of foucalt
testing that is inescapable:

The user has to judge the *brightness* of two *separated* patches of
light...and adjust the test apparatus so that the two patches are equal in
brightness (and ideally at 50% of max illumination).

Our eyeballs are not all that good at measuring brightness in absolute
terms, and are a bit better at relative measures (but with the foucalt test
the two areas we want to compare are sometimes not that close to one
another...making comparison more difficult and potentially less accurate).
(CCD/video camera based test rigs overcome much of this problem, but I'm
assuming we are making eyeball powered test rigs here.)

Our eyes judge *position* of shadows better and easier than *brightness* of
shadows.  Because of this, I recommend folks at least try the wire test and
compare its ease of use (and perhaps accuracy/repeatability) to the foucalt
test.

In many cases you don't have to make many changes to your test rig to
convert from foucalt to wire test.  (I use a pinhole light source, so all I
have to do is replace the knife edge with a wire and I'm off to the races.)

You may also find it easier to measure more zones with the wire test (using
a pin stick instead of a mask) than with foucalt.

Hope this helps,
Tom Krajci

PS.  I don't imply that we should discard foucalt testing.  I use both the
wire test and the foucalt test.  The foucalt test is a great way for me to
evaluate overall smoothness (no mask on the mirror), and check for hints of
turned edge.