Alain <ele### [at] netscapenet> wrote:
> Jim Holsenback nous illumina en ce 2008-12-23 08:32 -->
> > waves/ripples patterned density map applied to scattering media? ....
>
> A media won't distort anything.

Hah! Jim's idea is not so stupid after all; you're only correct as long as you
use "standard" Type 1 (isotropic) scattering. But if you use Type 2 (Mie haze)
or, probably better yet, Type 3 (Mie murky) scattering, this might actually get
you somewhere...

It's definitely not an exact model of what is really happening with hot air, but
still it's definitely worth a try.

> My take would be to place a long cylinder or a cone with pigment rgbt 1 give it
> an ior around 0.99 and use a slight normal perturbation: normal{bozo 0.1
> scale<100,1,1>} for a jet along the x axis. Rotate as needed.

Tried that. My experience is that if you pertube it strongly enough to get any
noticeable effect, you get ugly outlines due to total reflections.

Post a reply to this message