An RGB-meter cannot BY ITSELF tell if a subject is white
or gray, or put another way: the RGB-meter cannot measure the
reflectance of a subject. I don't know how it's supposed
to handle snow etc., but my guess is that it includes
contrast-analysis in setting the exposure.
One way to test this is the following setup: Put up a
stable light source (constant lamp, or non-TTL flash)
illuminate graycards of varying reflectance covering 5%
to 95%. Then put your F5 to work. My guess is it will
set different exposures for the 5% and the 95% cards.
Now put some ca. 18% reflectance subject in front of the
cards and repeat it. The exposures should now be better!
I do not have (access to) an F5, but it would be very
interesting if someone from the list would perform
this experiment.
Thomas Dall
