RE: fun with spheres...

problem solved...

I was not setting the VisDGeometry vertexFormat attribute properly.  I was
setting GeometryArray.NORMALS, but not GeometryArray.COORDINATES.  The
Java3D analoges to VisADGeometry seem insensitive to this...

Tom

> Hi Bill (and the rest of VisAD universe),
>
> I was able to fix one issue, I was mapping (index -> shape scale) and, go
> figure, my spheres were increasing in size over the 50k or samples in my
> data set...  oops.
>
> But the sphere/facet issue is still unresolved.  I have tested my geometry
> w/ a base line Java 3d setup and everything looks as desired...  Could the
> problem be my use of radial normals (based on the vertices) as opposed to
> polygon/face normals?
>
> Thanks much,
>
> Tom
>
> >
> >
> > Hi Tom,
> >
> > > I'm attempting to render points as spheres.  But i'm having
> problems w/
> > > rendering in POLYGON_FILL mode w/i visad.   My intention is to
> > have multiple
> > > solid, shaded spheres displayed of consistent size, but I get
> > spotty fill
> > > with inconsistent size...  Since a picture is worth a thousand
> > words i'll
> > > put links to pictures which best describe my situation.  The
> > second pic. is
> > > of the first zoomed into a sphere (which includes a shaded
> > come...???)  What
> > > am I missing?
> >
> > My best guess is that there is some problem with your
> > normals, or even errors in your geometry. The attached
> > jar file includes a Python program dna.py and a data
> > file dna_molecule.txt read by the program. This
> > program draws atoms as octahedra (poor man's spheres)
> > with no problem. Also, VisAD draws vary complex shapes
> > including spehres as iso-surfaces without problem.
> >
> > Good luck,
> > Bill
> >
>
>
>
>
>