Hmmmm! Looks like that rule of thumb formula I gave a bit earlier might be a bit out.(posted 8728 days ago)
I did some tests last night, taking readings from a lightbox surface using a photometer, and then from the film plane of a Nikon fitted with an f/1.2 lens, pointed at the same self-luminous surface.
It looks like the 20% figure given is a bit low. I was getting readings closer to 30%, and I tend to believe me, and my photometer, rather than some spurious figure off the web.
I now think the formula for film plane illumination should be:
Subject brightness x 0.3, divided by fnumber squared.BTW if anyone knows the correct, rather than empirical way to derive this, please let me know.