Christian Froeschlin <chr### [at] chrfrde> wrote:
> Not sure if this is related but ATM thickness of 300 and sun
> distance of 1000 seem a bit mismatched, although of course
> you can't use the real numbers due to scaling problems.

You're probably right; the only problem with reducing the size of the
bounding sphere further is that it becomes obvious and clashes with the
real horizon.

I've noticed that, when used together, the effect of the Mie scattering
media depends largely on the density of the Rayleigh media. After a certain
point, reducing the density of the former has no effect on the scene. For
now, I've replaced the media with following code:

  sphere {
    <0,0,0>, 1
    pigment { transmit 1 }
    hollow
    interior {
      media { // Rayleigh scattering
        scattering {4, <50,128,194>/255*0.5}
        density { rgb 0.05 }
        }
      }
    translate z*0.5
    }

  sphere {
    <0,0,0>, 1
    pigment { transmit 1 }
    hollow
    interior {
      media { // Mie murky scattering
        scattering {3, 0.05}
        density {rgb 0.05}
        }
      }
    translate z*-0.5
    }

and I've been playing around with the values. I can reduce the intensity of
the glare in the intersection of these two spheres, but not the size.

Post a reply to this message