Ron Parker <ron### [at] povrayorg> wrote:
> First, make sure the discs the lens effects plugin puts in front of
> the camera are specified as hollow.  That does make a difference,
> even for patch objects.  Unfortunately, it means you'll have to edit
> Chris's masterwork if they're not.

All lens effects are indeed set to be hollow and not to cast shadows,
precisely so the lens effects can be used with POV-Ray's various
atmospheric features.

> Second, you might need to add some transparent bounds to force the
> interval that passes through your scattering media to be shorter,
> or you may need to increase the number of intervals (as a last
> resort.)  I usually see this effect when the interval passing
> through the media is much larger than the media effect itself.  It
> is especially prominent when you specify atmospheric media.
>
> If your media is forward-scattering only, you can get good effects
> by adding a plane parallel to the camera plane just beyond the light
> source.  If it's backward-scattering only, add such a plane just
> before the light source and another one several units beyond the
> light source (depending on scale.)  If it's both, add a plane at
> the light source and another one beyond.  The idea is to force an
> interval to evaluate over the part of the media that has the most
> contribution to the scene.

As I replied to Ken, I suspect it may have something to do with the
distances involved in the creation of the lens effects...bounding the
media may well be part of the solution.

Post a reply to this message

Ken <tyl### [at] pacbellnet> demanded to know:
> How come your lens effects are attenuating my background scattering
> media into oblivion ?

To which Nieminen Juha <war### [at] punarastascstutfi> complacently replied:
>   Hard to say anything without seeing the code. Perhaps it has
something to
> do with max_trace_level? (Even an advanced user can fall in that
sometimes...
> I do :) ).

If you are using the pre-defined lens effects (Lens.inc) then this
should happen automatically - by default the file increases the level to
20.  With many layered effects or other transparent objects in the scene
this may not be enough, so the trick is to increase the level *after*
you include Lens.inc (otherwise the include file will set it back to
20).

Post a reply to this message

Ken wrote:
 
>  How come your lens effects are attenuating my background scattering
> media into oblivion ?

  The reason my media was not performing correctly had nothing to do with
the lens effects themselves. It turns out that I made a silly little mistake
when specifying light sources for the lens effects. You see if you don't
specify "media_interaction off" for the lens effects point light sources it
really messes with the scattering media in the scene. Once I changed that
everything started working as intended.

  It should also be noted that removing hollow from the lens effects disks does
really bad things and should be avoided.

  Just goes to show even advaced users can make mistakes and there is always
something new to learn with this splended program called POV-Ray.

-- 
Ken Tyler -  1200+ Povray, Graphics, 3D Rendering, and Raytracing Links:
http://home.pacbell.net/tylereng/index.html http://www.povray.org/links/

Post a reply to this message

you are the best, oncle Ken
Rodolphe
galerie: http://perso.infonie.fr/rod.gelaude
          annuaire "Povray Sites"
     http://www.multimania.com/povsite

news://news.newz.net/nzn.fr.3d.pov


>
>
>Ken wrote:
>
>>  How come your lens effects are attenuating my background scattering
>> media into oblivion ?
>
>  The reason my media was not performing correctly had nothing to do with
>the lens effects themselves. It turns out that I made a silly little
mistake
>when specifying light sources for the lens effects. You see if you don't
>specify "media_interaction off" for the lens effects point light sources it
>really messes with the scattering media in the scene. Once I changed that
>everything started working as intended.
>
>  It should also be noted that removing hollow from the lens effects disks
does
>really bad things and should be avoided.
>
>  Just goes to show even advaced users can make mistakes and there is
always
>something new to learn with this splended program called POV-Ray.
>
>--
>Ken Tyler -  1200+ Povray, Graphics, 3D Rendering, and Raytracing Links:
>http://home.pacbell.net/tylereng/index.html http://www.povray.org/links/

Post a reply to this message

Ken wrote:
> 
>   The reason my media was not performing correctly had nothing to do with
> the lens effects themselves. It turns out that I made a silly little mistake
> when specifying light sources for the lens effects. You see if you don't
> specify "media_interaction off" for the lens effects point light sources it
> really messes with the scattering media in the scene. Once I changed that
> everything started working as intended.
> 

Glad to hear you solved the problem. But I'm not sure I understand. You have two
light sources, in line with the camera, and the spot light in the background was
causing the media to display as white spots, right?
But why is it 'a silly mistake' to expect this spot light to work correctly with
scattering media? Why should the media display as white spots?

Margus

Post a reply to this message

Margus Ramst wrote:

> 
> Glad to hear you solved the problem. But I'm not sure I understand. You have two
> light sources, in line with the camera, and the spot light in the background was
> causing the media to display as white spots, right?
> But why is it 'a silly mistake' to expect this spot light to work correctly with
> scattering media? Why should the media display as white spots?
> 
> Margus

  Your guess is as good as mine. Media is a tricky beast and I will be the
first to admit that I am not a master with it.

  This should be enough for you to recreate it for yourself. For the two
point light sources turn media interation = on and you should see the
problem immediately.

camera{location<0,0,-.20>look_at<0,0,0>}

 light_source{< 0,.004,2>rgb<1,1,1>  media_interaction off}

 light_source{< 0,50,2>rgb<1,1,1>*.5 media_interaction off}

 light_source{<0,0,10>color rgb<.7,.3,.8> spotlight point_at<0,0,0>
 radius 1 falloff 10 tightness 1 media_interaction on}

media {
scattering {4, .1 extinction .25}
samples 5, 15
intervals 7
ratio .99
confidence .999
variance 1/100
}  

-- 
Ken Tyler -  1200+ Povray, Graphics, 3D Rendering, and Raytracing Links:
http://home.pacbell.net/tylereng/index.html http://www.povray.org/links/

Post a reply to this message

On Sun, 05 Dec 1999 14:05:34 -0800, Ken wrote:
>
>
>Margus Ramst wrote:
>
>> 
>> Glad to hear you solved the problem. But I'm not sure I understand. You have two
>> light sources, in line with the camera, and the spot light in the background was
>> causing the media to display as white spots, right?
>> But why is it 'a silly mistake' to expect this spot light to work correctly with
>> scattering media? Why should the media display as white spots?
>> 
>> Margus
>
>  Your guess is as good as mine. Media is a tricky beast and I will be the
>first to admit that I am not a master with it.
>
>  This should be enough for you to recreate it for yourself. For the two
>point light sources turn media interation = on and you should see the
>problem immediately.

The spotlight isn't necessary to get this effect.  The problem, as I had 
suspected, is that the sampling interval for the point lights is essentially
infinite, so the chance of getting a sample that scatters the light from the
point sources is vanishingly small.  If you add a backstop like

 plane {-z,-500 pigment {rgbt 1}}                     

to the scene, it renders as expected (though it's all white.)  Notice the 
lack of the hollow keyword.  Since I didn't expect significant contributions
from the media beyond 500 units, I elected to remove it from consideration and
speed up the render a tad.  (indeed, if you add hollow it's still all white,
as expected.)

-- 
These are my opinions.  I do NOT speak for the POV-Team.
The superpatch: http://www2.fwi.com/~parkerr/superpatch/
My other stuff: http://www2.fwi.com/~parkerr/traces.html

Post a reply to this message

Ron Parker wrote:
> 
> The spotlight isn't necessary to get this effect.  The problem, as I had
> suspected, is that the sampling interval for the point lights is essentially
> infinite, so the chance of getting a sample that scatters the light from the
> point sources is vanishingly small.

Ah! Now I see the light :)

Margus

Post a reply to this message

Ron Parker wrote:

> The spotlight isn't necessary to get this effect.  The problem, as I had
> suspected, is that the sampling interval for the point lights is essentially
> infinite, so the chance of getting a sample that scatters the light from the
> point sources is vanishingly small.  If you add a backstop like
> 
>  plane {-z,-500 pigment {rgbt 1}}
> 
> to the scene, it renders as expected (though it's all white.)  Notice the
> lack of the hollow keyword.  Since I didn't expect significant contributions
> from the media beyond 500 units, I elected to remove it from consideration and
> speed up the render a tad.  (indeed, if you add hollow it's still all white,
> as expected.)

The question that arises then is if this is desirable behavior. I expect
a point light to interact with media in a similar way as a spot light with
the exception that it is a uniform distribution of interaction rather than
a focused interaction like you recieve with a spotlight. To some degree
I understand why adding the plane works but wouldn't it be considered
a hack rather than desirable behavior ?

-- 
Ken Tyler -  1200+ Povray, Graphics, 3D Rendering, and Raytracing Links:
http://home.pacbell.net/tylereng/index.html http://www.povray.org/links/

Post a reply to this message

On Mon, 06 Dec 1999 10:01:11 -0800, Ken wrote:

>The question that arises then is if this is desirable behavior. I expect
>a point light to interact with media in a similar way as a spot light with
>the exception that it is a uniform distribution of interaction rather than
>a focused interaction like you recieve with a spotlight. To some degree
>I understand why adding the plane works but wouldn't it be considered
>a hack rather than desirable behavior ?

Yes, it is a hack.  Offhand, I can't think of a way to do the right thing
with respect to effectively infinite intervals, but I'll sleep on it.  
Suggestions are welcome, of course.  One idea I'm bouncing around is 
the notion that rays leaving the bounding sphere of the scene should be
split into intervals at the point where they leave the sphere, but I'm 
not sure what implications that would have.

-- 
These are my opinions.  I do NOT speak for the POV-Team.
The superpatch: http://www2.fwi.com/~parkerr/superpatch/
My other stuff: http://www2.fwi.com/~parkerr/traces.html

Post a reply to this message