Interesting. Approximately what I had in mind. Perhaps weighed averaging
might give a more uniform distribution though... Did I just reinvent an
oct-tree? ;)
I'll have to check if this actually works.
Thanks.

Margus

Post a reply to this message