I've redone the scene file found in POV-Ray\SCENES\ATMOS\atmo4.pov now that
I dug out the old 3.02 version and used it to compare it with 3.5 media. I
think I got it pretty close. You should be able to copy the atmosphere
declarations to another scene file and get similar renders to the old
'atmosphere' feature.

What I'm unsure of, is, how will it behave without a room to contain it? And
will there be artifacts due to differences in sampling, intervals and
whatever else. In other words, I only tweaked values until it looked okay.
Most notably is the lack of a 'distance' parameter for media, so I'm
guessing density might help with that. What is obvious in my edited scene
file is that I had to use low extinction values. I wished I didn't need to,
since that is always said to be a kind of artificial adjustment.

Oh yeah, I added the type 5 Henyey-Greenstein scattering into the file since
that was available in v3.02, yet wasn't inlcuded in the original atmo4.pov
example. Also, please keep the word "approximation" in mind.  ;-)

-- 
Bob H.
http://www.3digitaleyes.com

Post a reply to this message

Attachments:
Download 'Atmo35convert30_atmo4.pov.txt' (7 KB)

Hi Bob,
This much better. It looks very useful.
The lights interact nicely with the atmosphere.

Thanks

"Hughes, B." <omn### [at] charternet> wrote in message
news:3ffa4ac0@news.povray.org...
> I've redone the scene file found in POV-Ray\SCENES\ATMOS\atmo4.pov now
that
> I dug out the old 3.02 version and used it to compare it with 3.5 media. I
> think I got it pretty close. You should be able to copy the atmosphere
> declarations to another scene file and get similar renders to the old
> 'atmosphere' feature.
>
> What I'm unsure of, is, how will it behave without a room to contain it?
And
> will there be artifacts due to differences in sampling, intervals and
> whatever else. In other words, I only tweaked values until it looked okay.
> Most notably is the lack of a 'distance' parameter for media, so I'm
> guessing density might help with that. What is obvious in my edited scene
> file is that I had to use low extinction values. I wished I didn't need
to,
> since that is always said to be a kind of artificial adjustment.
>
> Oh yeah, I added the type 5 Henyey-Greenstein scattering into the file
since
> that was available in v3.02, yet wasn't inlcuded in the original atmo4.pov
> example. Also, please keep the word "approximation" in mind.  ;-)
>
> -- 
> Bob H.
> http://www.3digitaleyes.com
>
>
>

Post a reply to this message

"Khalid Hanna" <kha### [at] rogerscom> wrote in message
news:3ffab1ae$1@news.povray.org...
> This much better. It looks very useful.
> The lights interact nicely with the atmosphere.

Glad to hear it is better.

Everyone who compares 3.0 with 3.5, watch out for assumed_gamma global
setting. Most scenes created for version 3.0 (Windows) used a value of 2.2,
whereas 1.0 is more likely today in version 3.5.

If there were a purely numerical conversion of atmosphere (and halo) to
media this could be more exacting. For example, atmosphere 'distance' can't
be replaced solely with just media 'density'. I don't believe so anyhow. It
would likely need a spherical pattern centered on the camera and have a
fall-off, like that of fog, encompassing the realm as necessary (distance +
100% saturation = sphere boundary). All conjecture though, until tried.

Bob H.

Post a reply to this message

Hi Bob,
The file you supplied is Ok, but (there is always a but...) the spotlight
has a conic section and it discontinues shinig at a certain distance into
the scene (when I remove the Room box). I want to generate a light house
effect without having the dark misrable end of the cone to appear. I tried
many things but no luck.

Khalid

"Hughes, B." <omn### [at] charternet> wrote in message
news:3ffa4ac0@news.povray.org...
> I've redone the scene file found in POV-Ray\SCENES\ATMOS\atmo4.pov now
that
> I dug out the old 3.02 version and used it to compare it with 3.5 media. I
> think I got it pretty close. You should be able to copy the atmosphere
> declarations to another scene file and get similar renders to the old
> 'atmosphere' feature.
>
> What I'm unsure of, is, how will it behave without a room to contain it?
And
> will there be artifacts due to differences in sampling, intervals and
> whatever else. In other words, I only tweaked values until it looked okay.
> Most notably is the lack of a 'distance' parameter for media, so I'm
> guessing density might help with that. What is obvious in my edited scene
> file is that I had to use low extinction values. I wished I didn't need
to,
> since that is always said to be a kind of artificial adjustment.
>
> Oh yeah, I added the type 5 Henyey-Greenstein scattering into the file
since
> that was available in v3.02, yet wasn't inlcuded in the original atmo4.pov
> example. Also, please keep the word "approximation" in mind.  ;-)
>
> -- 
> Bob H.
> http://www.3digitaleyes.com
>
>
>

Post a reply to this message

"Khalid Hanna" <kha### [at] rogerscom> wrote in message
news:3ffb2d51$1@news.povray.org...
> Hi Bob,
> The file you supplied is Ok, but (there is always a but...) the spotlight
> has a conic section and it discontinues shinig at a certain distance into
> the scene (when I remove the Room box). I want to generate a light house
> effect without having the dark misrable end of the cone to appear. I tried
> many things but no luck.

This might necessitate using a media container. Testing for a way to let the
spotlight fade out correctly I could only do so by using the following in
the light source, and doing so within a much larger box (or sphere).

  fade_distance 10 // these values could vary, of course
  fade_power 2


I'm not able to prevent the cutting off (or oversaturating) of the end-cone
by adjusting samples higher (I went to 1000), adding more intervals
manually, changing ratio, confidence, variance, or even method type.

This is a typical difficulty when using global media. That is to say,
getting it to look right because it continues outward infinitely. There
might still be another answer, but I don't know of it.

-- 
Bob H.
http://www.3digitaleyes.com

Post a reply to this message