Hi Mike,

Adding +MB0 helped, thanks for the explaination!

Greetz,

JWV

"Mike Williams" <nos### [at] econymdemoncouk> wrote in message
news:xQW### [at] econymdemoncouk...
> Wasn't it JWV who wrote:
> >hello,
> >
> >30 minutes ago I started looking at the parametric object, and the result
> >was a bit strange (and slow). I used the piece of code from the help
files
> >to make a sphere, and then changed it to make a half sphere and gave it a
> >"pigment{White}". I was expecting to see a white half sphere, but what I
> >saw, was a blackish/grey-ish kind of shape.
> >This puzzeled me. So I made added a half white hollow sphere to the
scene,
> >to show you (the newsgroup) wat I was expecting, and what I was getting.
> >But then it even became more puzzling. There were now 2
> >half-white-hollow-spheres!
> >
> >Does anyone know what is happening here?
> >
> >The source will be posted in p.b.sf
>
> It looks like it might be a bounding bug, and may have something to do
> with shadow testing.
>
> The default Bounding Threshold is 3. [When there are only a few objects
> in a scene, the overheads involved in performing bounding tests are
> often not worth while, so POV doesn't perform bounding when there are
> less than this number of objects present.]
>
> When there were only two objects in your scene (the parametric object
> and the plane) bounding was disabled. Adding any third object caused the
> bounding code to be activated. [You can force bounding to activate by
> setting the Bounding Threshold to 0 with +MB0 in the command line.]
>
> There are other weird things going on. Changing your "contained_by" to a
> box causes the parametric to render correctly.
>
> I suspect the shadow testing code, since some of my isosurface tutorial
> examples have less than three objects in the scene and seem to render
> correctly (I'll check them all properly under 3.6 tomorrow). The thing
> that causes the difference in my example scenes is that my parametric
> examples have "no_shadow".
>
> -- 
> Mike Williams
> Gentleman of Leisure

Post a reply to this message