Not normals exactly but displacement mapping is something that does that.
The 'function' in isosurfaces can do it.  No idea why the normal statement
is not considered for such a thing as "deform" or displacement as you say;
makes sense to me.  Probably since it is a =faked= surface deformation
already and that would make it obsolete maybe? : )  In any case the change
would need the system used in the current patches using isosurface
functions.

Bob

"Jerry" <jer### [at] acusdedu> wrote in message
news:jerry-B58105.11121720121999@news.povray.org...
> I would like to see normals act the same as if they were 'real',
> (heightfields for flat objects). Besides just making more sense to me,
> it keeps open the possibility of a 'deform' keyword some day that
> actually does the deformation that the normal is pretending to do, with
> the obvious speed penalty.
>
> It may never happen, but it has been a dream of mine... to dream... the
> impossible dream...
>
> Jerry

Post a reply to this message