Thorsten Froehlich wrote:

> In article <3de879f4@news.povray.org> , Micha Riser <mri### [at] gmxnet> 
> wrote:
> 
>> Floating point error analysis is a bit more complicated than you seem to
>> think. For intersections calculations a ray-tracer needs to do a sequence
>> of calculations and the error which is produced by the limitation of
>> floating point precision is propagted. As there are different
>> calculations done for each primitives it is natural that the floating
>> point error effects are also different for each primitive. In short: What
>> you are seeing is definitely an effect of FP precision.
> 
> No, this is wrong.
> 
> There is a very simple answer and I am sure there is some FAQ about it
> somewhere as this question comes up every now and then.  The maximum
> distance for which an intersection is calculated in POV-Ray is limited. 

But if the max intersection distance (which is 10^7 according to frame.h) 
would be the issue here shouldn't all objects dissapear at the same 
distance then? And even if it is max distance that does it, the reason why 
there is a limit to the intersection distance is certainly to avoid 
floating point precision error.

- Micha

-- 
objects.povworld.org - The POV-Ray Objects Collection
book.povworld.org    - The POV-Ray Book Project

Post a reply to this message