Peter Popov wrote:
>It's the normal. Either the pattern (bumps, which uses different noise
>generators) or the non-uniform (anisotropic) scaling, which IIRC is
>handled differently in 3.5 due to some bugs in 3.1. Or maybe even
>both.

Just curious: in 3.5 the uniform scaling of a normal also scales the bump
height, so that the apparent slope stays the same, right? This far no
problem. But how does it work with non-uniform scaling, that is, how is the
height calculated?

Post a reply to this message