On Mon, 27 Mar 2000 12:49:56 +0100, Tom Melly wrote:
>
>Ron Parker <ron### [at] povrayorg> wrote in message
>news:38dbcb75$1@news.povray.org...
>
>> ..... Thus, most differences are
>> bounded by the bounding box of the first object in the difference (This
>does
>> not apply to differences that include inverted objects.)
>>
>> Thus, your difference is bounded by the box and the spheres are ignored.
>If
>> you can bound it more accurately, you should.
>>
>
>That's what I expected - ie. the differenced object can not be larger than
>the first object in the difference so apply that objects bounding. What's an
>inverted object?

One with the 'inverse' flag.  That is, 

  difference {
    sphere {0,1}
    sphere {x,1 inverse}
  }

is actually bounded by a tighter box than the one around the first sphere.
That's because it's actually an intersection of the two spheres, of course, 
but more complicated objects might have dozens of components in the difference
where only one or two are inversed like this.  In that case, the intersection
of all their bounding boxes is the bounding box of the object.  Also, you can
always rewrite such a difference unambiguously without the inverses:

  difference {
    object {A}
    object {B}
    object {C inverse}   
    object {D}
  }

can be rewritten as
  
  difference {
    intersection {
      object {A}
      object {C}
    }
    object {B}
    object {D}
  }



-- 
These are my opinions.  I do NOT speak for the POV-Team.
The superpatch: http://www2.fwi.com/~parkerr/superpatch/
My other stuff: http://www2.fwi.com/~parkerr/traces.html

Post a reply to this message