Hi,

QUICK QUESTION (kind of!):

Has anyone built the 'grey step / edge detector' optical illusion where
you are fooled into thinking that a piece of, for example, posterboard is
one uniform color, when in fact one half of the board is actually
distinctly darker than the other.  I have seen this illusion presented as
a rectangular board, mounted flat on a wall, with a "horse tail" (or
perhaps a long, thick bundle of yarn) hanging loosely from above the
middle of the board, so that it separates the board into two halves.  When
the horsetail is hanging down, separating the grey rectangle in half, it
looks like both sides of the board are exactly the same color, but when
you lift up the horsetail, you see that the two sides are actually quite
different shades of grey.  The horsetail prevents you from seeing the
sharp 'edge' that occurs where the two differently shaded sides of the
board meet, but does not act as an edge itself.  This illusion expolits
our inability to perceive difference in color value without some
sort of an edge or boundry.  It may not sound all that exciting, but the
effect is really quite impressive!

Anyway, I am working on building a version of this display for a
gallery demonstration on perception, and I've run into a technical detail
that I'm hoping someone out there in museum-L-land has already
encountered.  The illusion occurs only over limited differences in
greyscale shading.  If there is too big of a difference in the shade of
the two sides, we will be able to recognize that they are in fact two
different colors, even when we can't see the 'edge'.  Of course, the
illusion has the biggest impact when the difference is maximized.  Does
anyone know what percent grey shading tends to work best for the two
sides?  I'm currently experimenting, although unfortunately our computer
printer does not produce terribly even greyscale shadings, so the results
are difficult to interpret (and reproduce).  I've been told that this
illusion/exhibit is in one of the Exploratorium Cookbooks, but I do not
have access to any of them, and of course, we are pressed for time.

Thank you very much in advance for any help,

David