|
![](/i/fill.gif) |
In article <421cc4e9$1@news.povray.org>, "Slime" <fak### [at] email address>
wrote:
> > Have you considered writing the data out as a mesh2?
>
> I believe height fields trace considerably faster than their equivalent
> meshes.
Have you tried it? (I haven't...I'm really wondering if there's much of
a difference.)
A mesh has other advantages as well. You could create overhangs and
other structures, make it higher resolution tessellation in areas that
are more important, use a tessellation that doesn't produce artifacts
that are as obvious, etc. You can also compute the normals more
precisely while you're generating the mesh, rather than estimating them
from the triangle data. Look at the height field macros to see the
difference this can make...they compute the normal by looking at the
slope of the height function.
On the other hand, an image-based height field is easier to
edit...though you really need something that can handle 16 bit images.
--
Christopher James Huff <cja### [at] earthlink net>
http://home.earthlink.net/~cjameshuff/
POV-Ray TAG: <chr### [at] tag povray org>
http://tag.povray.org/
Post a reply to this message
|
![](/i/fill.gif) |