"Kenneth" <kdw### [at] gmailcom> wrote:

>
> Getting a dense, self-shadowing, *almost* pure-color media is a bit difficult.
> I'll probably post something about that separately. (This test is with
> scattering media-- plus extinction-- along with absorption media.)

Hmm, with further experimentation, it wasn't so difficult after all-- after I
worked around some of the pitfalls of using multiple medias (which warrant some
comments that I'll eventually post separately.)

A 'better' PURE-green cloud is the code below.

Note the 'extinction' value of 0. Although scattering media's extinction is good
for producing self-shadowing, it presently can't take a color vector, just a
float value. As-is, it creates a *complementary* color in the media (nearly
PURPLE in this case.) That's realistic, of course-- a colored cloud *should*
filter out lighting of that color-- but for artistic reasons like a 'pure'-color
cloud, it would be nice to be able to play around with extinction. The
additional absorption media below-- having NO color-- produces what looks like
identical self-shadowing, but without that complementary coloring.

-----
(needs at least one light source)

sphere{0,1
    hollow
        texture{
            pigment{rgbt 1}
            finish{ambient 0 emission 0 diffuse 0}
            }
        interior{
            media{
                 method 3
                 intervals 1
                 samples 70

                 scattering{1, rgb 50*<0,1,0> extinction 0} // NO extinction,
// as it introduces an unwanted 'complementary' color into the media's
// self-shadows.

                 absorption 50 // NO color

                 emission 8*<0,1,0> // just to lighten-up the self-shadows
// a bit

                 density{
                    spherical
                    color_map{
                        [.12 rgb 0]
                        [.37 rgb 1]
                             }
                        scale 3
                        warp{turbulence .7 omega .5}
                        scale 1/3
                       }
       translate 1.2*y
       }

--------
Play around with the medias' multiplier values, to get a density that looks
nice.

Post a reply to this message