Re: ATM Bartelized CCD Robo-Foucault success!

["Bob May" <bobmay@nethere.com>]

Yep, you have it right.  You should be looking at the gray, not the black
for when you are there.  The reason is that the black has one direction of
brightness (you can only go white) and the sensitivity is low at the black
from both the imager gain and the low amount of light getting through.  With
the gray, you get both the comparison between the two relative to each other
and a better sensitivty to the brightness of the image.  I'm not to sure,
but I would suspect that either 1/2 or 3/4 of the light impinging on the KE
would be where you want to set the brightness.  I am unsure as it's either
sensitivity goes up as the percentage of light getting cutoff per step of
cutting or not.  Worknig with one of the zones as black really doesn't
indicate anywhere near as well what the difference is between the two sides
as you could have one come out 1 brightness step and you will see a small
error and you then need to go and step the KE again sideways to find that
you still have an error as you then get another 1 or 2 counts in the image
brightness difference.  Better to get the difference from the gray and then
you know how far off they are.
Bob May
http://nav.to/bobmay
bobmay@nethere.com