Re: ATM Colorful Foucault test?

["Bill T." <twentiethwave@hotmail.com>]

Charles W. Grant wrote:
> The human eye is much more sensitive to intensity variation than to hue
> variation. Especially for detecting and comparing subtle gradients. You
> would need a very high contrast, low signal-to-noise ratio image to see
> a variation between red to violet to blue.  More likely it would be
> subtile pastels. Color variation would probably only be useful for
> computerized analysis.

That color sensitivity issue is technically true, but there is a practical
exception when density is color coded.  I was thinking of the following
case: when I view a full mirror Foucault photograph in shades of gray, it is
extremely difficult to trace the equal density null ring around its full
circumference, or even locate its equal density points on opposite sides --
that's the main reason for Couder masks.  If the same image is posterized so
that each density is encoded with a unique color, the null ring jumps right
out.  This very cool page sort of illustrates that:

http://home.netcom.com/~mpeck1/astro/autof.htm

If the color contrast were high enough (you point out that this may be
unlikely) it might be possible to resurrect a pin block type test where the
full mirror is always viewed, but instead of trying to locate extinction
points by density, we locate center of null by color contrast.

So, ahem, who is going pick up Virgil's pioneering experiments?

Bill Tondreau