Rick [Kitty5] <ric### [at] kitty5com> wrote:
: all ~50mb of it as exported from rhino :)

: takes about 1min to parse, and is much faster then using an iso surface,
: anybody wants the rhino 1.1 file email me.

  I made a mesh-version of that object created with #while-loops. The source
takes less than 1kb and parsing takes less than 30 seconds (in an average
computer). And you can easily control the resolution of the mesh by modifying
one single value at the beginning of the source.
  However, I got a problem which was too difficult. Although the vertices
were exactly where they should (according to the surface function), I
couldn't get the normals working. I didn't realize how to take into
account the twisting of the "sine-wave-radius cylinder" when calculating
the normal vectors. They were too steep in the inner side of the "torus"
and too wide in the outer side. The error was too visible in order for it
to be usable.
  Of course the correct normals can be calculated, but I got tired.
  Perhaps someone else wants to to think about this?

-- 
#local D=array[6]{11117333955,7382340,3358,3900569407,970,4254934330}
#local I=0;#macro M()<mod(D[I],13)-6,mod(div(D[I],13),8)-3,10>#end
#while(I<6)cylinder{M()#local D[I]=div(D[I],104);M().1
pigment{rgb M()}}#local I=(D[I]>99?I:I+1);#end              /*- Warp -*/

Post a reply to this message