I wonder how make rendering sphere_sweep faster.
I tried to add a bounded_by clause without any improvement.

Any Idea ?

Regards,

Gilles

---
http://lordikc.free.fr

Post a reply to this message

gbe54 nous apporta ses lumieres en ce 2007/09/13 15:43:
> I wonder how make rendering sphere_sweep faster.
> I tried to add a bounded_by clause without any improvement.
> 
> Any Idea ?
> 
> Regards,
> 
> Gilles
> 
> ---
> http://lordikc.free.fr
> 
> 
> 
The problem, is usualy a sparcely filled bounding box. This especialy true when 
you have large sweeping curves forming a narrow tube. Each ray that enter that 
bounding box have to be tested against the sweep itself, demanding that the 
whole sweep be evaluated for an intersection with that ray.
A possible solution is to subdivide the sphere_sweep into several smaler 
sphere_sweep. This way, you have many smaler almost full bounding box, instead 
of one BIG almost empty box.

-- 
Alain
-------------------------------------------------
Believe those who are seeking the truth; doubt those who find it.

Post a reply to this message

you can add an hollow statement as well ..









web.46e992d757a3c828589a27270@news.povray.org...
>
> I wonder how make rendering sphere_sweep faster.
> I tried to add a bounded_by clause without any improvement.
>
> Any Idea ?
>
> Regards,
>
> Gilles
>
> ---
> http://lordikc.free.fr
>
>
>

Post a reply to this message

Penelope20k <pen### [at] caramailfr> wrote:
> you can add an hollow statement as well ..

  Once again, "hollow" is a catch-all keyword to solve all problems,
the panacea of all keywords: It cures all problems. Or at least you
could try to use it to see if it cures the problem.

  Please don't take this the wrong way, I'm just honestly curious:
What is it with the "hollow" keyword that causes this phenomenon in
your case? You are not the first and certainly not the last to have
this misconception about that keyword. I have seen it suggested in
the weirdest of situations (which have absolutely nothing to do with
what "hollow" does).

  One of the weirdest situations where "hollow" was suggested as a
cure-all remedy for all problems which I remember was when someone
was asking why he was getting rendering artifacts with an isosurface.

  I have the strong opinion that the "hollow" keyword should be deprecated,
if for nothing else, to remove its status as a "panacea" keyword which
is suggested to solve all problems. It's very undescriptive as well
(because its only effect is that it allows media and fog inside the
object, something which could be represented with a much more descriptive
keyword).

-- 
                                                          - Warp

Post a reply to this message

Warp <war### [at] tagpovrayorg> wrote:
> Penelope20k <pen### [at] caramailfr> wrote:
> > you can add an hollow statement as well ..
>
>   Once again, "hollow" is a catch-all keyword to solve all problems,
> the panacea of all keywords: It cures all problems. Or at least you
> could try to use it to see if it cures the problem.
>

Why not change its name to Snake Oil?
I make everything hollow. I read that was what to do, in some chicken
entrails. :-)


Stephen

Post a reply to this message

Depending on your requirements, perhaps a chain of cones + spheres would work as a
substitute, and may render faster?  I realize this is not a direct answer to your
question, but may be a good
workaround.

Post a reply to this message

Kyle <hob### [at] gatenet> wrote:
> Depending on your requirements, perhaps a chain of cones + spheres would work as a
substitute, and may render faster?  I realize this is not a direct answer to your
question, but may be a good
> workaround.

  I wonder if nobody has made some include file which creates a mesh
from data similar to the one which you use to define a sphere sweep.
The mesh could render much faster while still looking the same.

-- 
                                                          - Warp

Post a reply to this message

Warp wrote:
> Kyle <hob### [at] gatenet> wrote:
>> Depending on your requirements, perhaps a chain of cones + spheres would work as a
substitute, and may render faster?  I realize this is not a direct answer to your
question, but may be a good
>> workaround.
> 
>   I wonder if nobody has made some include file which creates a mesh
> from data similar to the one which you use to define a sphere sweep.
> The mesh could render much faster while still looking the same.
> 

Mike Williams SweepSpline.inc is just wonderful... I made some nice snakes 
out of it.

--
Jaime

Post a reply to this message