Experimenting using parametric surfaces I noticed they render very
slowly. This is also the case for Micha Riser's very interesting
implementation of n-sided Bezier patches.
While preparing an animation concerning Dini's surface (for the result
see p.b.animations)I noticed that besides the very slow rendering an
important part of the surface (a peak) was cut off. So I returned to a
method, in an adopted form, dating from the 3.0 period as used and
thoroughly explained by Tore Nordstand
http://www.uib.no/People/nfytn/mathgal.htm . This method also constructs
smooth triangles and when using functions (but not with "parametric")
the source needs little adaptation for other parametric surfaces.
But my principal question remains: why do parametric surfaces render so
slowly? 
-- 
Herman Serras
Gent (Belgium)
http://cage.rug.ac.be/~hs/

Post a reply to this message