Food for thought:

 Ok we know that a sky sphere is a virtual effect and that for all intensive
purposes it is not a real object. On the same token it is not really a pigment
either because it cannot be applied to an object. This leave us with the
following observations that seem somewhat undefined as far as it's usage
(at least from the simple description in the docs)...


Attempting to render the following produces an error from Pov stating that
there are no objects in the scene...

--
camera{location z*-5 look_at 0}
sky_sphere{pigment{gradient y pigment_map{[0 blue 1][1 red 1]}}}
--

So it is not an object is it ?



Rendering the following with the addition of a box allows the render to
go without error because of the addition of an object to the scene...

--
camera{location z*-5 look_at 0}
box{-.1,.1 pigment{rgb 1}}
sky_sphere{pigment{gradient y pigment_map{[0 blue 1][1 red 1]}}}
--

So it is an object only when there is a real object in the scene ?
What if I only want to raytrace a sky_sphere for a background image ?




The next little ambiguity that we can try is the following...

--
camera{location z*-5 look_at 0}
box{-.1,.1 pigment{rgb 1}}

#declare Sky =
sky_sphere{pigment{gradient y pigment_map{[0 blue 1][1 red 1]}}}

object{Sky}
--

Whoops ! Now we get an error that an object is expected but a sky_sphere is
found instead. Now we know for sure that Pov does not recongnize it as an
object.



So we change things around again to accomodate the parser and try the
following code...

--
camera{location z*-5 look_at 0}
box{-.1,.1 pigment{rgb 1}}

#declare Sky =
sky_sphere{pigment{gradient y pigment_map{[0 blue 1][1 red 1]}}}

sky_sphere{Sky}
--

Well that does work but the question is WHY ? Because a sky_sphere is a sky_sphere
is a sky_sphere ! 


Which leaves us with the still unanswered question: What is a sky_sphere ?


Enjoy your meal,

-- 
Ken Tyler

See my 850+ Povray and 3D Rendering and Raytracing Links at:
http://home.pacbell.net/tylereng/index.html

Post a reply to this message

Ken wrote:
> 
> Food for thought:
> 
...
>
> Which leaves us with the still unanswered question: What is a sky_sphere ?
> 
> Enjoy your meal, 

(Dominee Gremdaat?)

> 
> --
> Ken Tyler
> 

Perhaps the same goes for background.
Anyway, your dealing here with the metaphysics of POV.
Another interesting question might be: what's BEHIND a sky_sphere (or a
background).
Not a galaxy-include...

Dominee Remdaat

Post a reply to this message

Ken <tyl### [at] pacbellnet> schreef in berichtnieuws
> Attempting to render the following produces an error from Pov stating
that there are no objects in the scene...
>
> --
> camera{location z*-5 look_at 0}
> sky_sphere{pigment{gradient y pigment_map{[0 blue 1][1 red 1]}}}
> --

And adding a lightsource to this small scene makes POV-Ray crash :(

Ingo

Post a reply to this message

ingo wrote:
> 
> Ken <tyl### [at] pacbellnet> schreef in berichtnieuws
> > Attempting to render the following produces an error from Pov stating
> that there are no objects in the scene...
> >
> > --
> > camera{location z*-5 look_at 0}
> > sky_sphere{pigment{gradient y pigment_map{[0 blue 1][1 red 1]}}}
> > --
> 
> And adding a lightsource to this small scene makes POV-Ray crash :(
> 
> Ingo

Can you provide an example so that I can confirm it ?

-- 
Ken Tyler

See my 850+ Povray and 3D Rendering and Raytracing Links at:
http://home.pacbell.net/tylereng/index.html

Post a reply to this message

Ken <tyl### [at] pacbellnet> schreef in berichtnieuws > Can you provide an
example so that I can confirm it ?

camera{location z*-5 look_at 0}
sky_sphere{pigment{gradient y pigment_map{[0 blue 1][1 red 1]}}}
light_source {<500,500,-500> rgb 1}


PII 233, NT4 SP4, POV3.1g msvc

Ingo

Post a reply to this message

ingo wrote:
> 
> Ken <tyl### [at] pacbellnet> schreef in berichtnieuws > Can you provide an
> example so that I can confirm it ?
> 
> camera{location z*-5 look_at 0}
> sky_sphere{pigment{gradient y pigment_map{[0 blue 1][1 red 1]}}}
> light_source {<500,500,-500> rgb 1}
> 
> PII 233, NT4 SP4, POV3.1g msvc
> 
> Ingo

It does not crash on my system

P200 mmx, Win 98, Pov3.1g.watcom

-- 
Ken Tyler

See my 850+ Povray and 3D Rendering and Raytracing Links at:
http://home.pacbell.net/tylereng/index.html

Post a reply to this message

Doesn't the usual concept of a 'sky_sphere' make it a background
pigment? Same as 'background' yet capable of color_map inclusion? So,
if you can figure background out you will have your answer perhaps 8-0
that all points not occupied by a object, and behind every object,
there is the sky sphere. Not a entity per se, moreso a environment.

Bob

Remco de Korte <rem### [at] xs4allnl> wrote in message
news:37C7DD9D.FD641BC8@xs4all.nl...
> Ken wrote:
> >
> > Food for thought:
> >
> ...
> >
> > Which leaves us with the still unanswered question: What is a
sky_sphere ?
> >
> > Enjoy your meal,
>
> (Dominee Gremdaat?)
>
> >
> > --
> > Ken Tyler
> >
>
> Perhaps the same goes for background.
> Anyway, your dealing here with the metaphysics of POV.
> Another interesting question might be: what's BEHIND a sky_sphere
(or a
> background).
> Not a galaxy-include...
>
> Dominee Remdaat
>

Post a reply to this message

Ken wrote:
> 
> Which leaves us with the still unanswered question: What is a
> sky_sphere ?

The sky_sphere is simply a way of adding a layer to the background.

According to the docs, you can legally define both a sky_sphere and
a background, and where the sky_sphere is transparent, the background
will show through.

The docs list the sky_sphere under atmoshperic effects, and not objects.

Regards,
John
-- 
ICQ: 46085459

Post a reply to this message

Remco de Korte <rem### [at] xs4allnl> wrote:
: Another interesting question might be: what's BEHIND a sky_sphere (or a
: background).

  That certainly is an interesting question. Perhaps if you move the
camera very very very very far away from the origin, you will reach the
border of the sky_sphere...

-- 
main(i,_){for(_?--i,main(i+2,"FhhQHFIJD|FQTITFN]zRFHhhTBFHhhTBFysdB"[i]
):5;i&&_>1;printf("%s",_-70?_&1?"[]":" ":(_=0,"\n")),_/=2);} /*- Warp -*/

Post a reply to this message

Nieminen Juha wrote:
> 
> Remco de Korte <rem### [at] xs4allnl> wrote:
> : Another interesting question might be: what's BEHIND a sky_sphere (or a
> : background).
> 
>   That certainly is an interesting question. Perhaps if you move the
> camera very very very very far away from the origin, you will reach the
> border of the sky_sphere...

That's no good because then you would fall off the edge of the world. Duh !

-- 
Ken Tyler

See my 850+ Povray and 3D Rendering and Raytracing Links at:
http://home.pacbell.net/tylereng/index.html

Post a reply to this message

Ken <tyl### [at] pacbellnet> wrote:
: That's no good because then you would fall off the edge of the world. Duh !

  But that would make a really good animation!

-- 
main(i,_){for(_?--i,main(i+2,"FhhQHFIJD|FQTITFN]zRFHhhTBFHhhTBFysdB"[i]
):5;i&&_>1;printf("%s",_-70?_&1?"[]":" ":(_=0,"\n")),_/=2);} /*- Warp -*/

Post a reply to this message

But it's impossible to move beyond the sky_sphere because it's infinite. It's
merely rendered as a sor of "backdrop" to the scene, the same regardless of
camera position. Kind of like those skies in Doom and Quake that don't appear
to get closer as you move.

Of course, that doesn't mean there's nothing else there... Suppose we live in
a three-dimensional universe that's like a "slice" of a four-dimensional one.
There's a whole nother dimension that we can't possibly get to.

Nieminen Juha wrote:

> Remco de Korte <rem### [at] xs4allnl> wrote:
> : Another interesting question might be: what's BEHIND a sky_sphere (or a
> : background).
>
>   That certainly is an interesting question. Perhaps if you move the
> camera very very very very far away from the origin, you will reach the
> border of the sky_sphere...
>
> --
> main(i,_){for(_?--i,main(i+2,"FhhQHFIJD|FQTITFN]zRFHhhTBFHhhTBFysdB"[i]
> ):5;i&&_>1;printf("%s",_-70?_&1?"[]":" ":(_=0,"\n")),_/=2);} /*- Warp -*/

Post a reply to this message

Larry Fontaine wrote:
> 
> But it's impossible to move beyond the sky_sphere because it's infinite. It's
> merely rendered as a sor of "backdrop" to the scene, the same regardless of
> camera position. Kind of like those skies in Doom and Quake that don't appear
> to get closer as you move.
> 
> Of course, that doesn't mean there's nothing else there... Suppose we live in
> a three-dimensional universe that's like a "slice" of a four-dimensional one.
> There's a whole nother dimension that we can't possibly get to.

  I think that you will find that the people in this thread know all about
how the sky sphere is implemented. The sky sphere is not even really infinite.
It exsists everywhere something else isn't. More specificaly everywhere there
is no object the skysphere fills in the unused pixels. A fancy fog of sorts.

We were just having fun...

-- 
Ken Tyler

See my 850+ Povray and 3D Rendering and Raytracing Links at:
http://home.pacbell.net/tylereng/index.html

Post a reply to this message

Ken <tyl### [at] pacbellnet> wrote:
:   I think that you will find that the people in this thread know all about
: how the sky sphere is implemented. The sky sphere is not even really infinite.

  Moreover: The sky sphere is actually a sphere of radius 1 with its
center at the camera location. The differences between a regular sphere and
the sky sphere are, AFAIK:
  1) Ray-object intersection tests are done with the sky sphere only if the
ray in question doesn't hit any other object. If the ray hits another object,
the sky sphere is completely ignored.
  2) The sky sphere is not shaded, that is, no lighting calculations are done
to the sky sphere but the color of the intersection point is taken as is.
  3) You can only apply a pigment to the sky sphere. This is a direct
consecuence of the previous feature (finish and normal are only applicable
when calculating shading; also reflection is not applicable since we are
handling a sphere of radius 1, which probably doesn't surround the whole
scene).

: We were just having fun...

  Yes, pov-humor, you know... Only povrayers understand pov-humor... :)

-- 
main(i,_){for(_?--i,main(i+2,"FhhQHFIJD|FQTITFN]zRFHhhTBFHhhTBFysdB"[i]
):5;i&&_>1;printf("%s",_-70?_&1?"[]":" ":(_=0,"\n")),_/=2);} /*- Warp -*/

Post a reply to this message

I have a question for the hardcore povrayers (I can only guess the correct
answer, but I'm not sure about it):

  The sky sphere is a sphere of radius 1 around the camera, as I described
in the previous article.
  Now suppose that the camera is at the origin and a ray is shot from it in
the z direction and it hits a reflective object at <0,0,2>. The reflected
ray is shot in the x direction (ie. it made a 90 degrees turn). This
reflected ray doesn't hit anything else.
  Since it doesn't hit anything else, it is tested against the sky sphere.
But since the sky sphere is a sphere of radius 1 at the origin, this ray
doesn't (?) hit it.
  How is this solved?

-- 
main(i,_){for(_?--i,main(i+2,"FhhQHFIJD|FQTITFN]zRFHhhTBFHhhTBFysdB"[i]
):5;i&&_>1;printf("%s",_-70?_&1?"[]":" ":(_=0,"\n")),_/=2);} /*- Warp -*/

Post a reply to this message

On 3 Sep 1999 07:26:01 -0400, Nieminen Juha wrote:
>  I have a question for the hardcore povrayers (I can only guess the correct
>answer, but I'm not sure about it):
>
>  The sky sphere is a sphere of radius 1 around the camera, as I described
>in the previous article.

Not true.  The sky sphere is a bit of code that takes a ray that didn't 
hit anything else and colors it based on its direction, according to the
documentation.  That fact invalidates the rest of your question. :)

Post a reply to this message

(Speculation)
I suspect the sky_spere is not exactly represented as an unit sphere around the
camera, but around the start of a ray (camera, reflected, refracted...) More
exactly, the returned value would be that of the pigment 1 unit away from the
origin, in the (relative) direction of the ray. Since only the direction is
important, this would give the desired result.

Margus

Nieminen Juha wrote:
> 
>   I have a question for the hardcore povrayers (I can only guess the correct
> answer, but I'm not sure about it):
> 
>   The sky sphere is a sphere of radius 1 around the camera, as I described
> in the previous article.
>   Now suppose that the camera is at the origin and a ray is shot from it in
> the z direction and it hits a reflective object at <0,0,2>. The reflected
> ray is shot in the x direction (ie. it made a 90 degrees turn). This
> reflected ray doesn't hit anything else.
>   Since it doesn't hit anything else, it is tested against the sky sphere.
> But since the sky sphere is a sphere of radius 1 at the origin, this ray
> doesn't (?) hit it.
>   How is this solved?
> 
> --
> main(i,_){for(_?--i,main(i+2,"FhhQHFIJD|FQTITFN]zRFHhhTBFHhhTBFysdB"[i]
> ):5;i&&_>1;printf("%s",_-70?_&1?"[]":" ":(_=0,"\n")),_/=2);} /*- Warp -*/

Post a reply to this message