I don't have control over media at all. It's just trial and error work here!
And there are tons of spheres, just a few of em ar visible. I use a small
perl script for creating the distributions. the media 3 uses both scattering
and absorption with 150 samples
