I was playing around with the brick pattern yesterday and came across
what I consider an anomaly: the "bricks" are not continuous, that is,
they do not exist in the center of the solid, but are "decals" that
exist only on the surface. To see what I mean, render a solid with a
brick pattern and make the mortar transparent. Instead of seeing the
individual "bricks", you see two flat rectangles on either side of the
solid.

camera {
  location  <-20.0, 5.0, -100.0>
  direction 4*z
  right     4/3*x
  look_at   <0.0, 0.0, 0.0>
}

light_source { <30, 30, -30> color rgb 1.0 }

box { -0.5, 0.5
  scale <200, 100, 3>
  texture {
    pigment {
      brick
        pigment { rgbt 1 },
        pigment { rgb <0.6, 0.1, 0.1> }
        mortar 1.0
    }
  }
}


 -----------------------------------------
|               Tim Riley                 |
|            Boulder, Colorado            |
|      E-mail:tri### [at] itsbldrdocgov      |
|     Per favore spenga le Sue scarpe     |
|  davanti abbandonare il nave spaziale.  |
 -----------------------------------------

Post a reply to this message

Tim Riley wrote in message <352CF3D7.E62B8CC0@its.bldrdoc.gov>...
>I was playing around with the brick pattern yesterday and came across
>what I consider an anomaly: the "bricks" are not continuous, that is,
>they do not exist in the center of the solid, but are "decals" that
>exist only on the surface. To see what I mean, render a solid with a
>brick pattern and make the mortar transparent. Instead of seeing the
>individual "bricks", you see two flat rectangles on either side of the
>solid.


As I see it, this is true for the other textures too, isn't it? I just tried
it out with a marble texture. As I see it, POV-Ray only calculates the color
of an object when the ray hits the objects surface. If this color is
transparent, then the ray continues through the object, but the next
hit-test is only made when it hits the next surface, in this case the
backside of the object.
What you would want, is for POV-Ray to also calculate the texture inside the
object again, probably at a specified sample rate.
That would be a great thing (my first thought is an onion-pattern like
texture with transparent outer-layers and a little turbulence added - this
in shades of grey and applied to a sphere would give you a great rock!).
Maybe the "interior" of the upcoming 3.1 version is something like that...

Bye, Johannes.

Post a reply to this message

On Thu, 09 Apr 1998 10:14:15 -0600, Tim Riley <tri### [at] itsbldrdocgov>
wrote:

>I was playing around with the brick pattern yesterday and came across
>what I consider an anomaly: the "bricks" are not continuous, that is,
>they do not exist in the center of the solid, but are "decals" that
>exist only on the surface. To see what I mean, render a solid with a
>brick pattern and make the mortar transparent. Instead of seeing the
>individual "bricks", you see two flat rectangles on either side of the
>solid.

All textures, except halos, are like this.  All objects are treated as
hollow "shells", so when a texture is transparent, the ray passes
right through it and doesn't get checked again until it hits another
object, or another side of the same object.  To do otherwise would be
computationally prohibitive.

Post a reply to this message

Ronald L. Parker wrote:
> All textures, except halos, are like this.  All objects are treated as
> hollow "shells", so when a texture is transparent, the ray passes
> right through it and doesn't get checked again until it hits another
> object, or another side of the same object.  To do otherwise would be
> computationally prohibitive.

Makes sense. Thanks.

 -----------------------------------------
|               Tim Riley                 |
|            Boulder, Colorado            |
|      E-mail:tri### [at] itsbldrdocgov      |
|     Per favore spenga le Sue scarpe     |
|  davanti abbandonare il nave spaziale.  |
 -----------------------------------------

Post a reply to this message