It was just experimenting.

I fixed the problem: I have to specify 'media_interaction off' for my
light_source. By doing this the white part of the beam before the lens (not
refracted yet) disappears. Apparently, no photons are deposited in this area
if the light has not been refracted (or reflected, I guess), I can't see
exactly why. Thus, I had to put a thin cylinder of glass with 'target' and
'refraction on' for photons in front of the light_source.

cylinder
{
  0, 0.01*z, 1.2*LIGHT_RADIUS
  hollow
  pigment {rgbt 1}
  interior {ior 1.0}
  translate -5*z
  photons {target collect off refraction on reflection on}
}

What this experiment tends to show is that photons are deposited in media
only if the light has been already refracted or reflected at least once
before reaching the current location. Am I right?

But this does not explain the artifact in the first case (with
'media_interaction on'). Could it be an 'extinction' side-effect of the
media though I specified zero?

I also noticed that, when using media photons and in the sorting phase of
the parse, the status bar displays a wrong number of sorted photons.


I join here another experiment. I am quite happy with the path of the red
beam in the first lens. However, I wonder why there is a gap between the
incident rays and relected rays on the mirrors. Suggestions welcome. And
there is something strange at the end where the remaning beams converge.

Post a reply to this message

Attachments:
Download 'experiment.png' (139 KB)

Preview of image 'experiment.png'