Mike Metheny wrote:
> 
> I think the texture difference; while perhaps unintended, looks very cool :>

Heh! :)

I was so focused on the problem that I did not see that!

Thank you.


Tor Olav
--
mailto:tor### [at] hotmailcom
http://www.crosswinds.net/~tok/tokrays.html

Post a reply to this message

"Tor Olav Kristensen" <tor### [at] hotmailcom> wrote in message
news:3900C930.F7039ED4@hotmail.com...
|
| I have now tried to replace the difference with an
| intersection of a sphere and a torus with the
| 'inverse' keyword. (Is this what you meant?)
|
| ==> Same result. No change in the reflections. :(

Nope.  I was talking about the torus surface itself, maybe if it's inversed then
perhaps the reflection will right itself by then being on the "outside".
Realize though that by turning the torus inside-out then the CSG will
change too.  I think that a inverse used in the intersection or difference
should cancel that change.
Rather than try and make guesses I just tried to make such a thing myself
and I couldn't even see any reversed (or whatever) reflection:

light_source { <-5,10,-10> color rgb 1.5
}
camera {
  location  <-1,1,-10>
  angle 25
  look_at   <0, 0, 0>
}
sphere {0,1
        pigment {gradient y color_map {
                [0,1 color rgb <.9,.2,.1> color rgb <.2,.4,.8>]
        }} scale 30
}

#declare Connector=
union {
difference { // torus core, capped
        sphere {0,1}
        torus {1.5,1}
      translate -y*.6615 // approximation only
}
difference { // torus anti-core
        torus {1.5,1}
        sphere {0,1 inverse}
      translate y*.6615 // approximation only
}
        pigment {rgb <.25,1,.25>}
        finish {reflection .5}
}

object {Connector
         scale <.5,2,.5> rotate <-30,70,-90> translate -y
}
object {Connector
         scale <.5,2,.5> rotate <30,-70,90> translate y
}

Post a reply to this message

Bob Hughes wrote:
> 
> Rather than try and make guesses I just tried to make such a thing myself
> and I couldn't even see any reversed (or whatever) reflection:

First: Thank you very much for trying to help me!

I'm quite frustrated, because I need the reflections from this 
middle "cigar" shape to behave correctly(?) in other scenes.

I have now included some of my source code below so that you can see 
what I'm actually doing.

Note that my convex "cigar" shape is made with a torus that has a 
greater minor radius than major radius.

And I now wonder if this could cause the odd change in reflections?

(As I understand from your code, you are making a similar shape by 
scaling a sphere along an axis.)

Regards

Tor Olav
--
mailto:tor### [at] hotmailcom
http://www.crosswinds.net/~tok/tokrays.html

// ===== 1 ======= 2 ======= 3 ======= 4 ======= 5 ======= 6 ======= 7

#version 3.1;

#include "colors.inc"
#include "textures.inc"

// ===== 1 ======= 2 ======= 3 ======= 4 ======= 5 ======= 6 ======= 7

#declare v3Origo = <0, 0, 0>;

#macro Pythagoras(hyp, kat)

 sqrt(hyp*hyp - kat*kat)

#end // macro Pythagoras

// ===== 1 ======= 2 ======= 3 ======= 4 ======= 5 ======= 6 ======= 7

#declare RadW = 1.0; // Max Radius for Wide   part (Cigar Shape)
#declare RadN = 0.5; // Min Radius for Narrow part (Spoke shape)

#declare RadMinC = 26; // Minor Radius for Cigar shape
#declare RadMinS = 26; // Minor Radius for Spoke shape

#declare HoleClearing = 0.1;

// ===== 1 ======= 2 ======= 3 ======= 4 ======= 5 ======= 6 ======= 7

#declare RadMajC = RadMinC - RadW; // Major Radius for Cigar shape
#declare RadMajS = RadMinS + RadN; // Major Radius for Spoke shape

#declare ss = Pythagoras(RadMinC + RadMinS, RadMajC + RadMajS);
#declare tt = ss*RadMinS/(RadMinC + RadMinS);

#declare DistBetween = 2*ss; // RadW + RadN

#declare Cigar =
difference {
  sphere { v3Origo, Pythagoras(RadMinC, RadMajC) }
  torus { RadMajC, RadMinC }
  cylinder { (RadW + RadN)*x, -(RadW + RadN)*x, RadN + HoleClearing }
//  plane {  y, tt - ss }
//  plane { -y, tt - ss }
}

#declare Spoke =
difference {
  sphere { v3Origo, vlength(<RadMajS, RadMinS>) }
  torus { RadMajS, RadMinS }
  plane {  y, -tt }
  plane { -y, -tt }
}

//#declare SpkCig =
merge { 
  object { Cigar translate +ss/2*y }
  object { Spoke translate -ss/2*y }
  texture { Soft_Silver }
  pigment { color Red }
}

// ===== 1 ======= 2 ======= 3 ======= 4 ======= 5 ======= 6 ======= 7

Post a reply to this message

I checked out what you had there and also didn't see any way of getting the inside to
become outside, which is the case for that torus with the larger minor radius than
major.
Not that it was going to be possible using inverse or variations thereof (I even gave
scale -1 a try), but as it turns out I couldn't even get the MegaPov
'interior_texture' to
put it right.
This means to me that something goes wrong with that torus.  Not a complete surprise
either since it's been said before by other people how that particular shape isn't
reliable.
The problem is in the inner part of the torus, even though usually an outer surface,
ends
up inside itself.  So the raytracer must be confusing it entirely.

Bob

Post a reply to this message

Bob Hughes wrote:
>
> This means to me that something goes wrong with that torus.  Not a complete surprise
> either since it's been said before by other people how that particular shape isn't
> reliable.
> The problem is in the inner part of the torus, even though usually an outer surface,
ends
> up inside itself.  So the raytracer must be confusing it entirely.

So it is the "ciagar" shape that messes up the 
reflections. 

That's a pity. =(

Is this to be considered as a bug ? 

And if so, do you know if the other people you 
mentioned have posted a bug report on the subject?

I have also noticed that this shape is hollow. 
And therefore it doesn't behave in CSG operations.

Tor Olav
--
mailto:tor### [at] hotmailcom
http://www.crosswinds.net/~tok/tokrays.html

Post a reply to this message

I don't know if it's a mathematic problem with the torus and if not then it must be
related to the raytracer.
Maybe there's a possible fix but I'm not one who could even begin to guess the answer.
I think it might have been said to be a "bug", but again I couldn't say.  Straining to
remember
I think people were conceding it to be a problem with the raytraced surface and might
not
be
easily fixed.  Takes those math geniuses out there (and here) I'm sure.

Bob

"Tor Olav Kristensen" <tor### [at] hotmailcom> wrote in message
news:3904F7A6.AF260EC3@hotmail.com...
|
| Bob Hughes wrote:
| >
| > This means to me that something goes wrong with that torus.  Not a complete
surprise
| > either since it's been said before by other people how that particular shape isn't
| > reliable.
| > The problem is in the inner part of the torus, even though usually an outer
surface,
ends
| > up inside itself.  So the raytracer must be confusing it entirely.
|
| So it is the "ciagar" shape that messes up the
| reflections.
|
| That's a pity. =(
|
| Is this to be considered as a bug ?
|
| And if so, do you know if the other people you
| mentioned have posted a bug report on the subject?
|
| I have also noticed that this shape is hollow.
| And therefore it doesn't behave in CSG operations.
|
| Tor Olav
| --
| mailto:tor### [at] hotmailcom
| http://www.crosswinds.net/~tok/tokrays.html

Post a reply to this message

In article <3904F7A6.AF260EC3@hotmail.com>, Tor Olav Kristensen 
<tor### [at] hotmailcom> wrote:

> So it is the "ciagar" shape that messes up the 
> reflections. 
> 
> That's a pity. =(
> 
> Is this to be considered as a bug ? 
> 
> And if so, do you know if the other people you 
> mentioned have posted a bug report on the subject?
> 
> I have also noticed that this shape is hollow. 
> And therefore it doesn't behave in CSG operations.

I think the reason you end up with a "cigar" shape at all is a bug. As I 
recall, this is because the bounding gets messed up by the radius 
settings, I think someone was working on a fix for it...which wouldn't 
give a cigar shape, but the actual torus shape.

-- 
Christopher James Huff - Personal e-mail: chr### [at] yahoocom
TAG(Technical Assistance Group) e-mail: chr### [at] tagpovrayorg
Personal Web page: http://chrishuff.dhs.org/
TAG Web page: http://tag.povray.org/

Post a reply to this message

Chris Huff wrote:
> 
> I think the reason you end up with a "cigar" shape at all is a bug. 

As far as I remember you do get the torus shape if you do not 
subtract it from another shape.

Btw: Are there other shapes that intersects with themselves 
like this? 

(Maybe one could introduce a new "self_intersection" keyword ? :)

> As I recall, 
> this is because the bounding gets messed up by the radius
> settings, I think someone was working on a fix for it...which wouldn't
> give a cigar shape, but the actual torus shape.

But, but, but ... 
...How are we supposed to model such cigar shapes then?

(Scaling a sphere along an axis gives a shape that's round 
at the ends, while this cigar shape ends in a sharp point.)

Tor Olav
--
mailto:tor### [at] hotmailcom
http://www.crosswinds.net/~tok/tokrays.html

Post a reply to this message

Tor Olav Kristensen wrote:

> But, but, but ...
> ...How are we supposed to model such cigar shapes then?

A lathe can do it with ease and can also be used in CSG operations.

-- 
Ken Tyler - 1400+ POV-Ray, Graphics, 3D Rendering, and Raytracing Links:
http://home.pacbell.net/tylereng/index.html http://www.povray.org/links/

Post a reply to this message

In article <390516B5.CA6D79B1@pacbell.net>, lin### [at] povrayorg 
wrote:

> Tor Olav Kristensen wrote:
> 
> > But, but, but ...
> > ...How are we supposed to model such cigar shapes then?
> 
> A lathe can do it with ease and can also be used in CSG operations.

You can also do it with the poly or quartic objects, and it is quite 
easy to do with the isosurface object.

-- 
Christopher James Huff - Personal e-mail: chr### [at] yahoocom
TAG(Technical Assistance Group) e-mail: chr### [at] tagpovrayorg
Personal Web page: http://chrishuff.dhs.org/
TAG Web page: http://tag.povray.org/

Post a reply to this message