Dear povray users,

I am fairly new to povray and am now working on a picture to illustrate
my scientific work. In this I have a block which I want to display as
transparent (it's a dielectric waveguide). Then I have a sphere_sweep
that sweeps a sin-curve to look as a ave traveling in the waveguide.
However when rendering this I get white dots of the boundary of the
sphere_sweep in the transparent block. Anyone have any ideas of how I
solve this?

The code for this can be found on: http://pastebin.com/f7b651d75 where
four files are pasted: render.pov, the main file; cnf.inc wich is not
part of my problem now; substrate.inc, which contains the substrate and
wave.inc which contains the objects waveN (N = 1, 2, 3...), which are
generated with GNU octave (a tool similar to Matlab), and are the sphere
sweeps. The transparent blocks are defined in render.pov and are found
on line 86 and 97 on pastebin.com.

Many thanks for any ideas!

Regards,
Robert Rehammar

Post a reply to this message

It would be helpful if you post the image so we can see the kind of white dots
you are talking about... or if you isolate the problematic elements in another
text file.

Post a reply to this message

Robert Rehammar wrote:
> 
> However when rendering this I get white dots of the boundary of the
> sphere_sweep in the transparent block. Anyone have any ideas of how I
> solve this?

This could be a coincident surfaces problem, as most little white (or 
black) dots are. Are the two surfaces overlapping each other exactly? If 
so, move one of the objects slightly and the dots should disappear.

Raytracers can never decide which object to render if both surfaces are 
fighting for the same rays.

I hope this solves the problem.

Sam

Post a reply to this message

ons 2008-08-06 klockan 12:27 -0700 skrev stbenge:
> Robert Rehammar wrote:
> > 
> > However when rendering this I get white dots of the boundary of the
> > sphere_sweep in the transparent block. Anyone have any ideas of how I
> > solve this?
> 
> This could be a coincident surfaces problem, as most little white (or 
> black) dots are. Are the two surfaces overlapping each other exactly? If 
> so, move one of the objects slightly and the dots should disappear.
> 
> Raytracers can never decide which object to render if both surfaces are 
> fighting for the same rays.
> 
> I hope this solves the problem.
> 
> Sam

Thanks Sam! This cleared the problem - the two surfaces were exactly as
you suspected coinciding in some places.

Regrds,
Robert

Post a reply to this message

Hi again,

I have a further problem down the line. That is, if I make the
sphere_sweep (semi) transparent then the white dots remain even when no
surfaces touch. That is, having one semitransparent object within
another gives problems on the boundary of the inner object. Anyone have
any experience/solutions with this?

Regards,
Robert

tor 2008-08-07 klockan 08:17 +0200 skrev Robert Rehammar:
> ons 2008-08-06 klockan 12:27 -0700 skrev stbenge:
> > Robert Rehammar wrote:
> > > 
> > > However when rendering this I get white dots of the boundary of the
> > > sphere_sweep in the transparent block. Anyone have any ideas of how I
> > > solve this?
> > 
> > This could be a coincident surfaces problem, as most little white (or 
> > black) dots are. Are the two surfaces overlapping each other exactly? If 
> > so, move one of the objects slightly and the dots should disappear.
> > 
> > Raytracers can never decide which object to render if both surfaces are 
> > fighting for the same rays.
> > 
> > I hope this solves the problem.
> > 
> > Sam
> 
> Thanks Sam! This cleared the problem - the two surfaces were exactly as
> you suspected coinciding in some places.
> 
> Regrds,
> Robert
>

Post a reply to this message

Robert Rehammar wrote:
> Hi again,
> 
> I have a further problem down the line. That is, if I make the
> sphere_sweep (semi) transparent then the white dots remain even when no
> surfaces touch. That is, having one semitransparent object within
> another gives problems on the boundary of the inner object. Anyone have
> any experience/solutions with this?

Are you using photons in your scene?  Recently I had issues with photons 
making white dots and I had to adjust the gather radius to deal with it. 
  According to the manual a radius/spacing ratio of 5:1 is recommended, 
so I used:

#declare Spc = 0.0025;
photons {
   spacing Spc
   radius Spc * 5
}

which worked perfectly.  Of course, the spacing value would depend on 
the scene.  Hope this is helpful.

-- 
-The Mildly Infamous Blue Herring

Post a reply to this message

Robert Rehammar wrote:
> Hi again,
> 
> I have a further problem down the line. That is, if I make the
> sphere_sweep (semi) transparent then the white dots remain even when no
> surfaces touch. That is, having one semitransparent object within
> another gives problems on the boundary of the inner object. Anyone have
> any experience/solutions with this?

Do you have radiosity turned on? Radiosity can cause strange circular 
artifacts from transparent objects...

Sam

Post a reply to this message

No, I do not have radiosity turned on, it's simply two transparent
objects within eachoter, the outer being a block and the inner a sphere
sweep.

Regards,
Robert

lör 2008-08-09 klockan 22:07 -0700 skrev stbenge:
> Robert Rehammar wrote:
> > Hi again,
> > 
> > I have a further problem down the line. That is, if I make the
> > sphere_sweep (semi) transparent then the white dots remain even when no
> > surfaces touch. That is, having one semitransparent object within
> > another gives problems on the boundary of the inner object. Anyone have
> > any experience/solutions with this?
> 
> Do you have radiosity turned on? Radiosity can cause strange circular 
> artifacts from transparent objects...
> 
> Sam

Post a reply to this message

> > > I have a further problem down the line. That is, if I make the
> > > sphere_sweep (semi) transparent then the white dots remain even when no

Try changing the sphere sweep into a simple merged while loop of spheres and
cylinders. Decrease the addition of X if you need higher resolution.
Hope this helps.
-Peter

camera{location z*-15 look_at 0}

light_source{<5,10,-20>,1}

merge{
    #local X = -6;
    #while (X<6)
        #local P = <X,sin(X),0>;
        sphere{P,0.25}
        #ifdef(P2)
            cylinder{P,P2,0.25}
        #end
        #local P2 = P;
    #local X = X+0.1;
    #end
pigment {rgb x*1 transmit 0.5}
}

Post a reply to this message