Gernot Frisch wrote:
> OUCH!!! What about a 16 bit desktop depth?
Render two p*****: One pass with the lower bits, one with the
upper bits :-P
One should use a PBuffer for such tasks anyway, and with a
PBuffer you're free to use 24 BPP also on a 16 BPP desktop.
> DO NOT DO THIS!
Why not? It's a standard technique.
> Best way.
> For speed issues, I'd check for a bounding sphere/ray collision
> first, this can be done very quickly and safes you the
> triangle-ray test for only a few triangles you would have to
> investigate further.
Hehe, what do you think OPCODE implements? Take a look into the
source code, the test hierachy is quite interesting.
Wolfgang Draxinger
--
E-Mail address works, Jabber: hexarith@[EMAIL PROTECTED]
ICQ: 134682867
|
