Well, it will do some of what the original poster would want, I would
think.  At least you get some light bouncing around with radiosity.

Interesting dillema though.  Am I the only one that when walking into a
well-lighted room starts thinking about all those rays bouncing around?
I think if everyone thought about it that question wouldn't be asked
quite so often.

-Mike

Ken wrote:

> It is a different process completely and will not achieve
> the properties the original poster was talking about.
>
> (The same yet different ?)
>
> Ken Tyler

Post a reply to this message

Dave Mann <dcm### [at] i-americanet> wrote:
: Does this mean no Disco Mirror Balls? No bouncing lasers off mirrors? No
: scattering light of broken mirror shards?

: Have I missed something very simple, or is the sad truth that light really
: can't be reflected off of reflective surfaces?

  You should read some documentation about raytracing, what you can do with
it and what you can't. Raytracing is not a perfect technique to model the
real world. I don't know if there is any.
  For example, with raytracing it's extremely difficult, if not impossible
to model area lights (area lights in povray are not true area lights as you
can read from the docs), reflective and refractive caustics, etc.
  You should think about how raytracing works and try to understand why
it's almost impossible to model caustics with it.
  There are other techniques to model area lights and caustics, for example
radiosity and photon mapping, but I don't know if they are very appliable
to a raytracing engine (ok, radiosity is, but I don't know about the
photon mapping).

-- 
main(i){char*_="BdsyFBThhHFBThhHFRz]NFTITQF|DJIFHQhhF";while(i=
*_++)for(;i>1;printf("%s",i-70?i&1?"[]":" ":(i=0,"\n")),i/=2);} /*- Warp. -*/

Post a reply to this message

It is not _almost_ impossible to create caustics with the standard
raytracing model - it's just impossible.
Both (true) radiosity and photon mapping should be applicable to a
raytracing engine. I am not sure if they can be applied to POV without
fundamental changes; I believe both would require some kind of surface
meshing.
POV's radiosity is not "true" radiosity - it uses a raytracing approach
to simulate radiosity effects and does no energy balancing of the scene.
I would love to have true caustics in POV, even if my scenes would take
days to render. They fundamentally increase the realism of many scenes
(and they look cool).

Margus

Nieminen Mika wrote:
 
>   You should read some documentation about raytracing, what you can do with
> it and what you can't. Raytracing is not a perfect technique to model the
> real world. I don't know if there is any.
>   For example, with raytracing it's extremely difficult, if not impossible
> to model area lights (area lights in povray are not true area lights as you
> can read from the docs), reflective and refractive caustics, etc.
>   You should think about how raytracing works and try to understand why
> it's almost impossible to model caustics with it.
>   There are other techniques to model area lights and caustics, for example
> radiosity and photon mapping, but I don't know if they are very appliable
> to a raytracing engine (ok, radiosity is, but I don't know about the
> photon mapping).
> 
> --
> main(i){char*_="BdsyFBThhHFBThhHFRz]NFTITQF|DJIFHQhhF";while(i=
> *_++)for(;i>1;printf("%s",i-70?i&1?"[]":" ":(i=0,"\n")),i/=2);} /*- Warp. -*/

Post a reply to this message

Margus Ramst wrote:

> POV's radiosity is not "true" radiosity - it uses a raytracing approach
> to simulate radiosity effects and does no energy balancing of the scene.

Not to mention that radiosity is a diffused reflection and
does not have incident angle reflection characteristics.

Ken Tyler

Post a reply to this message

Mike <Ama### [at] aolcom> wrote:
: Interesting dillema though.  Am I the only one that when walking into a
: well-lighted room starts thinking about all those rays bouncing around?

  Actually they are not rays, but waves.
  This may start a physics flame war, but a photon is not a particle but
a electromagnetic wave quantum (is that the right term? I have studied
physics only in finnish... :) ).
  Instead of thinking about the light as rays, think about it as if you
had a swimming pool full of water. When you make waves on the water,
the waves will bounce from the walls of the pool. At least I think about
it that way :)

-- 
main(i){char*_="BdsyFBThhHFBThhHFRz]NFTITQF|DJIFHQhhF";while(i=
*_++)for(;i>1;printf("%s",i-70?i&1?"[]":" ":(i=0,"\n")),i/=2);} /*- Warp. -*/

Post a reply to this message

I'd like to thank everyone who participated in this thread. Its been very
illuminating (but not reflective).

In less than 48 hours I've learn more about POV-Ray and its capabilities,
limitations, and functions because of the many web sites devoted to it and
these newsgroups. I've seen some astounding things and hope to contribute
something of value soon.

And yes... when I look around at the world now I am beginning to break
things down into its ray traced elements.

I'll have more questions I am sure. Thanks for answering this one!


-Dave

Post a reply to this message

On Tue, 01 Dec 1998 13:30:08 +0200, Margus Ramst <mar### [at] peakeduee>
wrote:

>Both (true) radiosity and photon mapping should be applicable to a
>raytracing engine. I am not sure if they can be applied to POV without
>fundamental changes; I believe both would require some kind of surface
>meshing.

Photon mapping doesn't require any meshing.  Whether it can be
smoothly added to POV is another question, but one which I hope to
find an answer to once I get the superpatch where I want it.

Post a reply to this message

>  Actually they are not rays, but waves.

Are you so sure? Most physics scientists aren't... They
agree that light sure behaves like a wave, but it also
behaves like a particle.

>  This may start a physics flame war, but a photon is not a
particle but
>a electromagnetic wave quantum (is that the right term? I
have studied
>physics only in finnish... :) ).

Well, yes , an "electromagnetic wave quantum" seems the best
way to visualise
(visualize ;) ) it. I guess the Finnish get the same physics
at school as in Holland... :)

Just wondering... does raytracing get any easier to
understand if you think about light as electromagnetic wave
quanta?

Julius Klatte
http://surf.to/jkhome

Post a reply to this message

Ah, physics concerning ray-tracing, good to hear. Bad to implement. Ha ha! 
that's a joke, get it? Oh well...
Seriously though, I'd vote for the old bent straws going from camera to 
surface to light source, but think of the clutter. Oh, I was going to be 
serious. I'm out of here.

Message <3665d600.0@news.povray.org>, Julius Klatte  typed...
>
>Just wondering... does raytracing get any easier to
>understand if you think about light as electromagnetic wave
>quanta?
>
>Julius Klatte
>http://surf.to/jkhome
>

-- 
 omniVERSE: beyond the universe
  http://members.aol.com/inversez/POVring.html
=Bob

Post a reply to this message

Ronald L. Parker wrote:

> Photon mapping doesn't require any meshing.  Whether it can be
> smoothly added to POV is another question, but one which I hope to
> find an answer to once I get the superpatch where I want it.

Photon mapping (i.e. Nathan's patched 3.1a source [1]) works for me.
(I had to remove a call to a function like "WIN_something" in tokenize.c
which didn't exist on my Unixen, and remove his C++-style comments
(real CCs don't like them)).

It is probably the thing Dave was looking for.
Rendering time increses by a factor of 10 ... 100.

I did not look at the superpatch yet - Nathan's page mentions it
as "old" in a section which itself is called "Old News", so I
decided to ignore it.

Ralf

[1]: http://nathan.kopp.com/patched.htm

Post a reply to this message