If your scale gets too large you may start getting visual effects
involving lots of black pixels and spots. The problem here is that
POV-Ray hard-codes a value for DEPTH_TOLERANCE which is different for
each shape varying from 1.0e-2 to 1.0e-8. This tolerance determines
whether an intersection point is close enough to the surface to be
considered on the surface. Without it, ray-object intersection
calculations could calculate an intersection point on the other side of
the surface and cause it to be a dark or bright spot.
As your scale increases, these numbers lose precision and eventually
become ineffective. You start getting ray trace acne or shadow acne.
I would say you're probably safe on the scale of 1000 or even 10,000,
but once you get to millions you may start seeing problems. The depth
tolerance for polygons is the smallest at 1.0e-8 so if you want to
experiment, polygons would be a good candidate.
On 2022-06-03 11:48 p.m., HackerDaGreat57 wrote:
> I'm pretty new to POV-Ray and have gotten somewhat experienced with its
> coordinate system and how it places cameras.
> I'm wondering what is the limit for coordinates? Take Blender, for example - it
> starts losing precision after a few thousand meters, but what about POV-Ray?
> TL;DR: What is the max reliable coordinate number in POV-Ray
Post a reply to this message