|
|
Hi,
I encountered a really strange behaviour of Povray 3.6 when doing csg with a
mesh2 object. As you can see in the attached image there are parts of the
objects and/or shadows missing.
The actual CSG operation is done with an infinite plane and the red
transparent boxes (looking like planes) are for reference only. The
Mesh-object can be found in scenes/objects/mesh2.pov.
Pseudo code is:
// CSG with mesh object
difference {
object { Mesh_A
texture { Mesh_TextureA }
}
plane { z,0 }
}
// reference cutting "plane"
box { -2,2
scale <1,1,.00001>
pigment { color rgb x transmit .9 }
no_shadow
}
Fig 1. When adding the transparent box all looks like it is supposed to
be athough it is hollow and not a solid object. Remember that the box is
actually no part of the CSG. The same look can be achieved when transmit
is set to 1 so that the box is invisible.
Fig 2. This is the actual csg object. Without the transparent box, the
object looks really strange. The inside of the object is missing in the
image of the object itself and in the reflection (notice that the missing
parts are different in the image and the reflection).
Fig 3. When cutting away the lower half of the object (plane { y,0 }), some
parts of the shadow disappear although the object itself looks ok (from
this viewpoint). When removing the "no_shadow" of the box the missing parts
mysteriously show up again (not shown in this image).
Is this a knwon bug of povray or am I just doing forbidden things, meaning
that CSG with mesh2 is not allowed (I did not find anything about this in
the manual). Is there another method of doing this (especially "solid" CSG
opposed to the "hollow" CSG of Fig 1)?
Florian
Post a reply to this message
Attachments:
Download 'mesh2-csg.jpg' (25 KB)
Preview of image 'mesh2-csg.jpg'
|
|
|
|
Zarniwoop wrote:
> Is this a knwon bug of povray or am I just doing forbidden things, meaning
> that CSG with mesh2 is not allowed (I did not find anything about this in
> the manual). Is there another method of doing this (especially "solid" CSG
> opposed to the "hollow" CSG of Fig 1)?
>
> Florian
>
Are you trying this with surface subdivision? That mesh object looks
*very* familiar :) If you are trying it with SS, you might want to take
the SS block out of the mesh2 statement and see if you still get strange
results.
The only mesh-csg I've tried so far has been with height_fields (which
work very well in csg operations).
-Sam
Post a reply to this message
|
|