David Fontaine wrote:

> So keep doing this process on the output

Yep

> - but how do you get that complexity?

Same way as the seirpinski triangle (I think that is how it is spelt)
or the menger sponge (Idea-idea-idea - media-media-media-df3-MegaPOV)

Think of it this way:
you start with an image, scale it down a little & rotate it a bit, mapping
onto a plane such that the camera is looking at the seam corner then paint a
transparent sphere on it & do the same to that image, resulting in each image
having one more sphere than before plus the amount added by viewing
overlapping images & if the scaling is below 0.5 relative to the camera fov
soon (10 frames) there are millions of spheres

> Outside of the transparent spheres it'd all be the same each time...

Unless you do one of the following (I did all 3)
1. you don't use "once" in the image_map{}
2. make sure that the camera overlaps two of the repeated images
3. transform (eg rotate) the image relative to the camera

BTW I can post the src if wanted
--
Bye
Pabs

Post a reply to this message