From: "Jerald F. Wright" 
To: ATM List 
Reply-To: "Jerald F. Wright" 




I'm  not sure what specific reflectivity you are looking to find.

Are you looking for reflectivity throughout the visual range?

I may be thinking wrong here but it seems to me to get an accurate determination
for some range such as visual range for example that you would need to take
readings at many wavelengths through the range.  The resulting plot of
reflectivity may be sufficiently linear that discrete R G B will give a
good result.  But I am not sure.

I have seen in a number of sources a number of something like 88% visual
reflectance for aluminum.  New of course. So your number could be quite
good if the aluminum is not new.  Seems to me the only way to know would be
a calibrated
light source.

It may be a good idea that rather than compare two reflective surfaces, use just
one surface and place a filter of known absorption in the light path and
calibrate by the known loss through the filter and without the filter.  Or
you might image your light source directly without reflection on to the ccd
and then
using the same imaging system (lens) reflect off the mirror. I would think
that it would be better to have the f number of your source to be higher
than the COC
f number of the mirror so that you know all of the light from the source is
reflected back into the ccd.  In fact a defocused image on the ccd may be a
way of comparing.  What size defocused image achieves the same pixel values
after reflection as some certain size defocused image size without
reflection.  Then the ratio of the number of pixels involved in each image
would "reflect" the reflectivity.

Another Idea may be to reflect the light source off an uncoated flat and
vary the angle of the reflection off that flat there by varying the
percentage of reflection.  Compare angles of reflection that achieve equal
pixel values after reflection off the surface in question and the
calibration is a physical property of the test and not to some possibly
arbitrary external reference.

Just a few ideas without having taken time to think of what could be wrong
with my thinking.

Jerry

--- BBBS/NT v4.00 MP