POV-Ray : Newsgroups : povray.binaries.images : Wall macro test : Re: Wall macro test Server Time
2 Nov 2024 18:50:48 EDT (-0400)
  Re: Wall macro test  
From: Bill Pragnell
Date: 26 Apr 2010 04:40:00
Message: <web.4bd550d45f2bb9c26dd25f0b0@news.povray.org>
"Nekar Xenos" <nek### [at] gmailcom> wrote:
> >   Isn't it extremely slow to render? How about making each brick a
> > similarly-generated heightfield instead? (I know that you can't achieve
> > the *exact* same level of detain with heightfields, but for bricks I
> > don't think it matters, unless you intend to look them from *really*
> > close.)
>
> The last time I tried doing large walls with height-fields, it was very
> slow and sometimes I couldn't render because of not enough memory. But
> then it was on a smaller computer than I have now. Also the bricks didn't
> look as nice...
>
> Currently I tried making simple bricks with triangles using the same
> method for placing the bricks with very interesting results:
> The triangles rendered slower than the isosurface bricks - about 30%
> slower! This was using 3.7 beta 36. I haven't tested on version 3.62

I wrote some mesh-weathering macros which can create bricks using the same
technique. Walls like this rendered with my macros render something like 10x
faster than using isosurfaces, although as you say there can be a large memory
hit if you want close-up detail. You only need to declar a small number of
different brick meshes in an array, however, then just pick, rotate and texture
them randomly when placing them.

This image contains less than 10 individual meshes for the stonework (not
including the statue):

http://news.povray.org/povray.binaries.images/thread/%3Cweb.49c00fd346c868fb219167190@news.povray.org%3E/?ttop=331089&t
off=300

Bill


Post a reply to this message

Copyright 2003-2023 Persistence of Vision Raytracer Pty. Ltd.