Hi,

I'm not sure if this is the right group, since the problem could by older. If I
have a union of simple CSG-objects (no differences, isosurfaces or other objects
known to produce incorrect values with min_extend or max_extend) containing a
light_source, I yield values near to infinity. A simple point light should
occupy no space at all IMO. An area light should be the box it is in. Has anyone
of you experienced the same? I could not find an answer with the search option
in the newsgroups.

Best regards,
Michael

Post a reply to this message

MichaelJF wrote:

> I'm not sure if this is the right group, since the problem could by older. If I
> have a union of simple CSG-objects (no differences, isosurfaces or other objects
> known to produce incorrect values with min_extend or max_extend) containing a
> light_source, I yield values near to infinity.

I can reproduce this using code like below. I'd consider this a
bug, although occuring in a rather rare combination of use cases.
My preference would be to ignore the light_source itself in extent
calculations as it is not geometry and has no effect on things like
trace function or object pattern / inside object test. Not sure about
looks_like objects (BTW these have no effect on current behavior).

#declare OBJ = union
{
   sphere {y,1 pigment {color White}}
   light_source {y color White}
}

#debug concat("***",str(max_extent(OBJ).x,1,1),"\n")

Post a reply to this message

Now that you mention it, I remember having had this same problem a 
couple of years ago. At the time, I did not relate it to the included 
light source, but that seems to make sense indeed. Not sure, but I think 
that the problem also exists with version 3.6.

Thomas

Post a reply to this message

Thank you both,

as you experienced the same I'm free of searching for mistakes within my code. I
was simply puzzled. I think it is not a bug but only a matter of interpretation.
A light source is infinite since it illuminates the whole scene. May be the
programmer of min_ and max_extend had this in his mind. I think, in this case
the light source adds one or more (area lights) points to the object under
investigation, simply the positions, which could be considered in calculating
the extend of the object. I don't think it should be ignored completely in this
case. But this is only another possible interpretation...


Best Regards,
Michael

Post a reply to this message