clipka <ano### [at] anonymousorg> wrote:
> Am 23.07.2012 22:52, schrieb Chaanakya:
> > clipka <ano### [at] anonymousorg> wrote:
> >> Am 23.07.2012 21:58, schrieb Chaanakya:
> >>> "Chaanakya" <nomail@nomail> wrote:
> >>>> Hey guys!  I just had a quick question...
> >>>>
> >>>> I was trying to generate the graph of this function (in Cartesian coordinates):
> >>>>
> >>>> -z + 1.00003 = (3x^2 + 3y^2)/(200000)
> >>>>
> >>>> However, when I render the following code, nothing shows up except for the
> >>>> plane:
> >> ...
> >>>> isosurface {
> >>>>     function {
> >>>>       -y - ((3*pow(x,2) + 3*pow(z,2))/(200000)) + 1.00003
> >>>>     }
> >>>>     // contained_by { box { -2,2 } }
> >>>>     pigment {
> >>>>       color Red
> >>>>     }
> >>>> }
> >> ...
> >>> Even more strangely, when I use the equally valid function
> >>>
> >>> function {
> >>>     y + 3*pow(x,2)/200000 + 3*pow(z,2)/200000 - 1.00003
> >>> }
> >>>
> >>> I get a cube.  I think there's something fundamental about isosurfaces that I'm
> >>> not understanding?  That is, how should I convert the function z = -3x^2/200000
> >>> - 3y^2/200000 + 1.00003 into an isosurface?
> >>
> >> Do un-comment the "contained_by" line!
> >>
> >> At x=0,z=0 you have y = 1.00003, which is outside the default
> >> contained_by object (box{1,1}), and even at the maximum x and z
> >> (x=1,z=1) you have y = 1.00000, which just barely touches the box.
> >>
> >> I.e. the surface you defined is (for practical purposes) all outside
> >> default container; so the inside of the default container is either
> >> completely outside the isosurface (first version) so that you don't see
> >> anything, or completely inside (second version) so that you simply see
> >> the container's shape.
> >
> > I'm trying to figure out exactly what container I should use - if I use box {
> > <-1,1,-1>,<1,1.00003,1> } I get nothing (I understand why).  If I use box {
> > <-1,0.9,-1>,<1,1.00003,1> } I get a box.  How do I get the elliptic paraboloid
> > to show up?
>
> At the current dimensions, the curvature of the isosurface is simply too
> small to be noticeable
>
> You'll need to use a much larger bounding container (and move back the
> camera a whole lot) to see anything. Something like
> contained_by{box{-20000,20000}}.
>
> (BTW your elliptic paraboloid seems pretty circular to me.)

I think I figured out another way of doing what I want to do using spheres.  I
will try it out and report back.

Post a reply to this message