"ajtribick" <nomail@nomail> wrote:

> Yes the testing I have been doing since I made this animation led me to the same
> conclusion. Has this been reported in an official bug reports forum?

Don't know.  While experimenting with this last year, I Google'ed the subject
and came across an *old* bug report (or message thread) that mentioned the same
thing. I assumed that the POV team was aware of it. They probably are; it must
be a difficult problem to solve (perhaps tied in with transparency issues.)

In looking back over my own POV test scene (and its MANY notes), I see that I
*did* come up with a workaround. An odd one, to be sure.  Try this: Make a
small sphere, out of camera view, fill it with media, then INVERSE the sphere.
According to my notes, that 'fills the universe' with media, and the scene no
longer shows those radiosity problems! (But to see the media, there needs to be
a 'background object' in this case, like a large flat box way off in the
distance, filling the frame, for the media to appear against. The box could be
textured with some kind of sky pigment or pattern, in your case.)

I'll do some tests again, to see if this is an 'always-true' workaround.

My notes also tell me that plain 'atmospheric' media--that is, not enclosed in
an object--works well too; no rad problems. I'll try that again as well.

KW

Post a reply to this message

"Kenneth" <kdw### [at] earthlinknet> wrote:
>  Try this: Make a
> small sphere, out of camera view, fill it with media, then INVERSE the sphere.
> According to my notes, that 'fills the universe' with media, and the scene no
> longer shows those radiosity problems!

Well...I can't seem to reproduce that result now. I don't remember how I came up
with that idea, but it doesn't work (although my old notes say it does.) Can't
trust those old notes. :-$

Sorry for sending you on a wild goose chase! I'll continue to try and figure out
my error...

Ken W.

Post a reply to this message

"ajtribick" <nomail@nomail> wrote in message 
news:web.48fb31f9ed540e8abc82ce420@news.povray.org...
> Test animation of a sunrise using media and radiosity. The atmosphere is 
> created
> using Rayleigh scattering media which has a density that falls off 
> exponentially
> with height. The test is only partially successful: it seems that the 
> media
> atmosphere is not giving any contribution to the radiosity lighting. Still 
> have
> to figure out if there's any way to increase the contribution of the 
> scattering
> media to the radiosity while still allowing the colour of the sky to 
> remain
> visible.

Just a thought... As you are having a problem getting the media to 
contribute to the radiosity of objects within the object containing the 
media, why not excavate a volume for the objects from that object as 
follows:

union
{
    difference
    {
        sphere{0, 100000}
        sphere{0, 1000}
        interior{media{...}}
    }
    union
    {
        //Other objects all within the inner sphere above
    }
}

Rarius

Post a reply to this message

Kenneth nous illumina en ce 2008-10-21 19:48 -->
> "Kenneth" <kdw### [at] earthlinknet> wrote:
>>  Try this: Make a
>> small sphere, out of camera view, fill it with media, then INVERSE the sphere.
>> According to my notes, that 'fills the universe' with media, and the scene no
>> longer shows those radiosity problems!
> 
> Well...I can't seem to reproduce that result now. I don't remember how I came up
> with that idea, but it doesn't work (although my old notes say it does.) Can't
> trust those old notes. :-$
> 
> Sorry for sending you on a wild goose chase! I'll continue to try and figure out
> my error...
> 
> Ken W.
> 
Filling an inverted sphere out of view is akind of filling the whole scene with 
uncontained media. You don't need the sphere, just the media definition.

-- 
Alain
-------------------------------------------------
When you can't tell the difference between satire and what your opposition 
actually believes, its about time to have a good laugh at it all

Post a reply to this message

Rarius nous illumina en ce 2008-10-23 04:45 -->
> "ajtribick" <nomail@nomail> wrote in message 
> news:web.48fb31f9ed540e8abc82ce420@news.povray.org...
>> Test animation of a sunrise using media and radiosity. The atmosphere is 
>> created
>> using Rayleigh scattering media which has a density that falls off 
>> exponentially
>> with height. The test is only partially successful: it seems that the 
>> media
>> atmosphere is not giving any contribution to the radiosity lighting. Still 
>> have
>> to figure out if there's any way to increase the contribution of the 
>> scattering
>> media to the radiosity while still allowing the colour of the sky to 
>> remain
>> visible.
> 
> Just a thought... As you are having a problem getting the media to 
> contribute to the radiosity of objects within the object containing the 
> media, why not excavate a volume for the objects from that object as 
> follows:
> 
> union
> {
>     difference
>     {
>         sphere{0, 100000}
>         sphere{0, 1000}
>         interior{media{...}}
>     }
>     union
>     {
>         //Other objects all within the inner sphere above
>     }
> }
> 
> Rarius 
> 
> 
Similar to, but more complicated than:

sphere{0,1000 interion{media...}}hollow texture{...}}

No need to use the union.

-- 
Alain
-------------------------------------------------
If you pick up a starving dog and make him prosperous, he will not bite you; 
that is the principal difference between a dog and a man.

Mark Twain

Post a reply to this message

Alain <ele### [at] netscapenet> wrote:
> Rarius nous illumina en ce 2008-10-23 04:45 -->
> > "ajtribick" <nomail@nomail> wrote in message
> > news:web.48fb31f9ed540e8abc82ce420@news.povray.org...
> >> Test animation of a sunrise using media and radiosity. The atmosphere is
> >> created
> >> using Rayleigh scattering media which has a density that falls off
> >> exponentially
> >> with height. The test is only partially successful: it seems that the
> >> media
> >> atmosphere is not giving any contribution to the radiosity lighting. Still
> >> have
> >> to figure out if there's any way to increase the contribution of the
> >> scattering
> >> media to the radiosity while still allowing the colour of the sky to
> >> remain
> >> visible.
> >
> > Just a thought... As you are having a problem getting the media to
> > contribute to the radiosity of objects within the object containing the
> > media, why not excavate a volume for the objects from that object as
> > follows:
> >
> > union
> > {
> >     difference
> >     {
> >         sphere{0, 100000}
> >         sphere{0, 1000}
> >         interior{media{...}}
> >     }
> >     union
> >     {
> >         //Other objects all within the inner sphere above
> >     }
> > }
> >
> > Rarius
> >
> >
> Similar to, but more complicated than:
>
> sphere{0,1000 interion{media...}}hollow texture{...}}
>
> No need to use the union.
>
> --
> Alain
> -------------------------------------------------
> If you pick up a starving dog and make him prosperous, he will not bite you;
> that is the principal difference between a dog and a man.
>
> Mark Twain

Thanks for the help everyone, I ended up using

difference {
 sphere { ... }
 sphere { ... }
 hollow
 interior { ... }
}

.... rest of scene ...

and the radiosity has worked. Here's the new version of the animation.

Post a reply to this message

Attachments:
Download 'mediaatmosphere2.mpg' (314 KB)

Alain <ele### [at] netscapenet> wrote:

> Filling an inverted sphere out of view is akind of filling the whole scene with
> uncontained media. You don't need the sphere, just the media definition.
>

Yes, that finally occurred to me--after posting my message, of course. :-S

But I still can't get that simple idea to work correctly. I'm just doing
something wrong, probably very obvious.

KW

Post a reply to this message

"Rarius" <rar### [at] rariuscouk> wrote:


> Just a thought... As you are having a problem getting the media to
> contribute to the radiosity of objects within the object containing the
> media, why not excavate a volume for the objects from that object as
> follows...

I actually never thought of doing that. Nice idea, very much appreciated!
Thanks.

Ken W.

Post a reply to this message

"Kenneth" <kdw### [at] earthlinknet> wrote:
> >  Try this: Make a small sphere, out of camera view, fill it with
> > media, then INVERSE the sphere. [And add a scene-filling object in the
> > background, for the media to appear against.]
> > According to my notes, that 'fills the universe' with media, and the scene
> > no longer shows those radiosity problems!
>
> Well...I can't seem to reproduce that result now. I don't
> remember how I came up with that idea, but it doesn't work (although
> my old notes say it does.) Can't trust those old notes. :-$
>
> Sorry for sending you on a wild goose chase! I'll continue to try and figure
> out my error...
>

For what it's worth, I did finally figure out how to reproduce this. Strangely,
it depends on the Bounding_Threshold value in the .ini file (or on the command
line.) By setting the threshold to one MORE than the total number of objects in
the scene, the inverse trick works, and the media appears everywhere outside the
sphere. (Lights don't count in this total.) If the threshold value is <= the
total number of objects, inverse has no effect, and the media still appears
inside the media object. Bounding_Threshold shouldn't behave that way, AFAIK.

I also found something else strange (if my eyes are not deceiving me): Using
inverse, the new 'universe-filling' media is not contributing anything to the
rad lighting. Only the INVISIBLE media in the sphere is (!), and again in the
same incorrect way. But there shouldn't be any media in the sphere now.

Perhaps there are hidden clues here to rad's odd behavior with object media.

Ken W.

Post a reply to this message

Kenneth nous illumina en ce 2008-11-05 23:00 -->
> "Kenneth" <kdw### [at] earthlinknet> wrote:
>>>  Try this: Make a small sphere, out of camera view, fill it with
>>> media, then INVERSE the sphere. [And add a scene-filling object in the
>>> background, for the media to appear against.]
>>> According to my notes, that 'fills the universe' with media, and the scene
>>> no longer shows those radiosity problems!
>> Well...I can't seem to reproduce that result now. I don't
>> remember how I came up with that idea, but it doesn't work (although
>> my old notes say it does.) Can't trust those old notes. :-$
>>
>> Sorry for sending you on a wild goose chase! I'll continue to try and figure
>> out my error...
>>
> 
> For what it's worth, I did finally figure out how to reproduce this. Strangely,
> it depends on the Bounding_Threshold value in the .ini file (or on the command
> line.) By setting the threshold to one MORE than the total number of objects in
> the scene, the inverse trick works, and the media appears everywhere outside the
> sphere. (Lights don't count in this total.) If the threshold value is <= the
> total number of objects, inverse has no effect, and the media still appears
> inside the media object. Bounding_Threshold shouldn't behave that way, AFAIK.
> 
> I also found something else strange (if my eyes are not deceiving me): Using
> inverse, the new 'universe-filling' media is not contributing anything to the
> rad lighting. Only the INVISIBLE media in the sphere is (!), and again in the
> same incorrect way. But there shouldn't be any media in the sphere now.
> 
> Perhaps there are hidden clues here to rad's odd behavior with object media.
> 
> Ken W.
> 
> 
> 
> 
> 
> 
If there is nothing between you and the background, you get "infinite" paths. 
Media is not eveluated along an infinite path. If you have a beam of light, 
then, things change, the samples are taken within the beam.
Where you have your inverted sphere, the path is no longer infinite, it stops at 
the sphere.

-- 
Alain
-------------------------------------------------
Adult, n.: One old enough to know better.

Post a reply to this message