Hi all, I'm doing a project about computer vision. my task is to reconstruct
the surface from a pair of images. the algorithm i am using bases on
helmholtz reciprocity. i generated some "real" images with Radiance to test
my program. but the intensity of pixels seems to be somewhat wrong. I use
plastic material and both specularity and roughness are zero. can someone
tell me whether the intensity of the pixels match the real world value or it
just looks real. Please forgive me for my poor English and the problem is
driving me mad. Any help will be appreciated.
roughness are zero. can someone tell me whether the intensity of the pixels match the real world value or it just looks real.
Should match real world, that is what most validation studies tell. Still, how do you interpret the pixel values, and how did you define material and light sources? Are you sure you did not get confused with units somewhere? It is hard to debug a scene we do not know, maybe you could describe the setup or even put a sketch to some public ftp or website?
Good luck, and I always thought that man mad(1) should be a redirect to man rpic(1) 
CU Lars.