I am attempting to simulate an x-ray using the material light attenuation
ability in POV-Ray. Unforutunatly I am getting an odd "aurora" around parts that
are intersecting one another, like this experiment:

http://img844.imageshack.us/i/22952805.png/

I want it to look more like this (one part is located behind the other instead
of intersecting):

http://img291.imageshack.us/i/99583176.png/

This intersection is unavoidible as my final images require hardware to be
placed inside of bones.  The models are all CAD models imported into POV-Ray
using Rhino. I was wondering if anyone had any solutions or suggestions for my
"aurora" issue.

Thanks.

Post a reply to this message

Alan <nomail@nomail> wrote:
> I am attempting to simulate an x-ray using the material light attenuation
> ability in POV-Ray. Unforutunatly I am getting an odd "aurora" around parts that
> are intersecting one another, like this experiment:

  I don't know if it's the reason, but I wonder if increasing max_trace_level
would fix the problem.

-- 
                                                          - Warp

Post a reply to this message

> I am attempting to simulate an x-ray using the material light attenuation
> ability in POV-Ray. Unforutunatly I am getting an odd "aurora" around parts that
> are intersecting one another, like this experiment:
>
> http://img844.imageshack.us/i/22952805.png/
>
> I want it to look more like this (one part is located behind the other instead
> of intersecting):
>
> http://img291.imageshack.us/i/99583176.png/
>
> This intersection is unavoidible as my final images require hardware to be
> placed inside of bones.  The models are all CAD models imported into POV-Ray
> using Rhino. I was wondering if anyone had any solutions or suggestions for my
> "aurora" issue.
>
> Thanks.
>
>

Try shifting one object by a small value, like 0.0001 unit. You may have 
some coincident surfaces and the small shifting will separate those. You 
may instead need to scale one object up or down by a small value like 
1.000001 or 0.999999.

As warp daid, try increasing max_trace_level.

You can also try scaling everything up by, say, a factor of 10 (model in 
mm instead of cm).


Try using a merge to remove internal surfaces, where that's acceptable. 
I suspect that it may not always be acceptable in your case, you 
possibly need those surfaces. But try it anyway for the various parts of 
a single object.



Alain

Post a reply to this message

Adjusting the max trace level didn't change the effect. It appears that surfaces
that run parallel to the light source causes the greatest discrepancies. It is
as if the ray runs along these surfaces and thus bypasses attenuation effects of
the material.

I have found that at lower surface transparency (.9 transmitance instead of 1),
the effect is greatly diminished. This causes other effects that aren't ideal
but the images could be used as necessary.

Post a reply to this message