i am trying to raytrace light (using pointlights or distant lights) on a
surface through a lens.
any refraction seen directly from the camera through the lens is ok.
but i get the feeling that light that passes through the lens and hits
an underlying surface is just the result of filtering the affected
surface with a (color) factor. i mean: no raytracing.
e.g. with real raytracing light through a convex lens would converge.
this is why you can start a fire with a looking glass in the sun. this
is why you get that strange pattern underwater (which needs extreme
amounts of time and power to raytrace of course) and the 'dancing
lights' in the shadow beneath your wineglass on the table .
concave lenses are not a problem since they disperse lightrays. the
raytraced result would usually just look like how lw is doing it right
am i right about the 'half-fake' raytracing (or am i imagining things)?
if i'm right, does anyone have any ideas to simulate the light
convergence through glasses (with/without liquid) or lenses?
* please remove KILLSPAM from the e-mail address when replying *
* please include 'William Li' in address *
van holsteijn en kemna BV
* research * design * engineering *
- Delft - Netherlands
- Brussels - Belgium