"Norbert Kern" <nor### [at] t-onlinede> wrote:

>
> The AO render uses good radiosity settings, no background, a simple
> texture and an emissive sphere like that:
>
> sphere {
>         0, 10000
>         texture {
>                 pigment {color rgb 1}
>                 finish {emission 1.2 diffuse 0}
>         }
> }
>
> #declare T2 =
> material {
>  texture {
>   pigment {color rgb 1}
>   finish {ambient 0 diffuse 0.8}
>  }
>  interior {ior 1}
> }
>

Yes, that make sense for the initial AO render. Thanks. Although, I'm not sure
what the reason is for interior{ior 1}.

My own 'fake AO' method-- actually just a few experiments years ago-- was to
have a similarly white object, but I also used a white plane under it, to pick
up the AO effect where the object met the plane. And a similar white outer
sphere--but used just as a background, and without radiosity(!). To create the
AO effect, I placed hundreds of point lights in a distant hemispherical pattern.
The many overlapping shadows on the object looked *similar* to a radiosity
render. But if I were to do it again, I would use your method ;-)

> Rune's Illusion code is used to create a pigment pattern with perfect fit
> for a texture map:
>
[snip]

Wow! I need to digest your code (your math trickery is something I need to
study; it uses techniques that I'm not too familiar with-- but that I really
need to learn and apply.) In all of my own uses of illusion.inc, I've never
actually tried 'mathematically' fitting the projected image to an object. I've
always done the opposite: choosing a pre-made digital photo image, then
constructing all the scene's geometry to 'fit' the photo details, when
'projected' onto the geometry! A very tedious procedure, I admit. Your code is
very much appreciated, as an alternative method for a different kind of use.

Post a reply to this message