Pov 3.5 RC6 on Intel 1.7Ghz 512M Ram running Windows 2000

I read in the help that "cubic splines produce the most flexible and smooth
curves" but this does not seem to be the case with sphere sweep. The
linear:spline and b_spline work as advertised but the cubic-spline seems to
curl in on itself and produce a sharp step. This code illustrates this. Is
it supposed to do this?

//pov 3.5 RC6
global_settings { assumed_gamma 1.1 }

#include "stdinc.inc"

  background { color Blue }
  camera {
    location <0, 2, -10>
    look_at  <0, 0,  0>
  }

light_source { <0, 2, -10> colour White }

  sphere_sweep {
    //linear_spline
    //b_spline
    cubic_spline
    9,
    <0, 10, 0>, 1
    <0, 10, 0>, 1
    <0, 2, 0>, 1
    <0,  1, 0>, 0.5
    <0,  0, 0>, 0.3
    <0, -1, 0>, 0.5
    <0,  -2, 0>, 1
    <0,  -10, 0>, 1
    <0,  -10, 0>, 1

    tolerance 0.000001 //0.2
     pigment { color Yellow }
 }

Richard

Post a reply to this message

On Mon, 10 Jun 2002 21:29:52 +0200, "Richard" <pop### [at] iprolinkch> wrote:
> I read in the help that "cubic splines produce the most flexible and smooth
> curves" but this does not seem to be the case with sphere sweep. The
> linear:spline and b_spline work as advertised but the cubic-spline seems to
> curl in on itself and produce a sharp step. This code illustrates this. Is
> it supposed to do this?

Personally I think that's your mistake. Whole sphere_sweep object is NOT
spline. It is object which controls its parameters with splines. The path
where centers of spheres are placed is spline. The radius function in 'time'
is spline. If the max radius is 'bigger' than curvature/run of center spline
then smaller radii with theirs centers are completly hidden inside big radius.
Moreover You took quoted sentence from 'lathe' chapter not from sphere_sweep
one. Lathe is not sphere_sweep either. I hope my answer was understable since
my English limits such explanations.

ABX

Post a reply to this message

Thanks for the detailed explanation. I know the center path is a spline as I
use this extensively in another scene where it works very well. I then went
on to assume that one could control the outer surface as well. It seemed a
reasonable assumption based on the linear spline and b spline results.Thats
the trouble with making assumptions. I will stick with the b spline - it
works fine for my needs.
Richard

news:9a8bguo643ubbncriskpc1a9pumoqtn6pe@4ax.com...
> On Mon, 10 Jun 2002 21:29:52 +0200, "Richard" <pop### [at] iprolinkch> wrote:
> > I read in the help that "cubic splines produce the most flexible and
smooth
> > curves" but this does not seem to be the case with sphere sweep. The
> > linear:spline and b_spline work as advertised but the cubic-spline seems
to
> > curl in on itself and produce a sharp step. This code illustrates this.
Is
> > it supposed to do this?
>
> Personally I think that's your mistake. Whole sphere_sweep object is NOT
> spline. It is object which controls its parameters with splines. The path
> where centers of spheres are placed is spline. The radius function in
'time'
> is spline. If the max radius is 'bigger' than curvature/run of center
spline
> then smaller radii with theirs centers are completly hidden inside big
radius.
> Moreover You took quoted sentence from 'lathe' chapter not from
sphere_sweep
> one. Lathe is not sphere_sweep either. I hope my answer was understable
since
> my English limits such explanations.
>
> ABX

Post a reply to this message

Thanks for the detailed explanation. I know the center path is a spline as I
use this extensively in another scene where it works very well. I then went
on to assume that one could control the outer surface as well. It seemed a
reasonable assumption based on the linear spline and b spline results.Thats
the trouble with making assumptions. I will stick with the b spline - it
works fine for my needs.
Richard

news:9a8bguo643ubbncriskpc1a9pumoqtn6pe@4ax.com...
> On Mon, 10 Jun 2002 21:29:52 +0200, "Richard" <pop### [at] iprolinkch> wrote:
> > I read in the help that "cubic splines produce the most flexible and
smooth
> > curves" but this does not seem to be the case with sphere sweep. The
> > linear:spline and b_spline work as advertised but the cubic-spline seems
to
> > curl in on itself and produce a sharp step. This code illustrates this.
Is
> > it supposed to do this?
>
> Personally I think that's your mistake. Whole sphere_sweep object is NOT
> spline. It is object which controls its parameters with splines. The path
> where centers of spheres are placed is spline. The radius function in
'time'
> is spline. If the max radius is 'bigger' than curvature/run of center
spline
> then smaller radii with theirs centers are completly hidden inside big
radius.
> Moreover You took quoted sentence from 'lathe' chapter not from
sphere_sweep
> one. Lathe is not sphere_sweep either. I hope my answer was understable
since
> my English limits such explanations.
>
> ABX

Post a reply to this message