|
|
POVMAN <s### [at] acom> wrote:
> By accident yesterday I discovered a neat trick. I used clipped_by /
> bounded_by on a CSG and this cut a hole , through the shape exposing the
> interior walls of the individual objects making up the CSG. Has anyone
> deliberately used this in an an image?
Well, that's precisely why clipped_by exists in the first place (and has
existed from at least pov2.2 and maybe earlier).
Basically clipped_by is an intersection of the object being clipped and
the clipping object, but the latter does not add any surface to the result
(unlike a true intersection CSG).
I believe that the reason clipped_by exists at all is that it's faster
to render than an intersection CSG. If you are cutting parts of an object
and it doesn't matter that the cuts are not covered by a surface (eg.
because the cuts are not seen in the image) then it's more efficient
to use clipped_by than intersection.
I wouldn't say there's any other useful usage of clipped_by, at least
not in realistic scenes (this is because seeing the inside of an object
which has been clipped actually means that you are seeing an infinitely
thin surface, which can't exist in real life). Of course you can use it
for mathematical visualization of surfaces, but that's not a common theme
in povray images.
bounded_by { clipped_by } is a handy shortcut to bound the object with
the same object that is clipping it. Also the reverse exists, which is
basically the same thing. Currently the importance of this has diminished
a lot because the bounding algorithms used in povray have been greatly
improved compared to what they were when bounded_by and clipped_by were
first introduced.
--
- Warp
Post a reply to this message
|
|
|
|
Warp <war### [at] tagpovrayorg> wrote:
> POVMAN <s### [at] acom> wrote:
>
> Well, that's precisely why clipped_by exists in the first place (and has
> existed from at least pov2.2 and maybe earlier).
Earlier, sort-of. DKBtrace didn't have clipped_by, but you could get the
same effect with its version of bounded_by.
Post a reply to this message
|
|