> not lossy jpeg please. PNG would be favourite. Please
> cut out only the bit that is interesting, to keep
> filesizes down.

png is a good choice... will be done..


> This is not actually changing the normals, you are just
> changing the direction of the lightsource relative to the
> model.

euh, you're correct of course. Blame my unfortuned choice of words as English is only my third language ;). The reason i asked is because i'm building up a list of light behavior for each normal index. Although i'm still far from done (2048 values to check manually for the XY plane only) i'm starting to wonder on how to appreach the next problem : the one of inclination and how the light on the normals reacts to it.

> Am tempted to write a program to examine all TS normals
> and grade them based upon the sides that they appear on-
> would expect to see some statistical hint of the encodings

> Problem is, I have no time :(
on the same boat here.. free time is limited...:-/