Re: [grt-talk] Some suggestions.

[Anton N. Mescheryakov]

Nikodemus Siivola wrote:

> On Thu, 8 May 2003, Anton N. Mescheryakov wrote:
>  
>
> Definitely.  We need a first-intersetion accelerator (or several).
>
> The only problem Z-buffer version is that Z-buffer algorythms for implicit
> surfaces tend to get comlicated pretty quickly. Not undoable, just more
> difficult for arbitrary surfaces. If you feel like up to it, ... ;)
>
> My current plan is to project the *bounding volumes* onto screen, and store
> the depth of the closest bounding surface per pixel. This will allow the
> tracing alogorith to step into correct point in octree without intermediate
> tests.
>  
The sad fact is that you can't bound an implicit surface in general 
case, without prior knowledge of its shape. Just about every implicit 
surface can turn out to be infinite, especialy isosurfaces; it looks 
like Povray handles infinite stuff separately from the rest. Full-scale 
analysis of implicit surface isn't something impossible (just some 
symbolyc derivatives) it's something too big for the current state of grt.
But finite primitives: quadrics, polyhedra etc are quite suitable for 
scanline rasterizers. *Even for hardware-accelerated*

> Alternatively we could store sorted lists of objects per screen region (per
> pixel or smallish group). Kind of like poor mans ray-sorting.
>  
IMHO, it's called A-buffer or so. KYRA series of 3D cards use (er... 
used) this stuff; it proved to be bandwidth saver... and display 
artifact source. And again, it can be hardware accelerated. Please don't 
think that I suggest you mold grt into some kind of realtime renderer, 
but 3D is now everywhere and it's a shame to ignore your second processor...


Best regards,
   Anton.