Here's some experimentation with use of radiosity to imply a "light
source over there". It uses a large inverted sphere with a gradient
pigment that is black to the left and white to the right.

I also have an idea about using a "radiosity light bulb"-- perhaps not
an original one.   When I tried it, I got just a few big white spots and
then only after putting ambient to like rgb 200 or something.   Has
anyone ever had success here;  any suggestions on which radiosity
parameters to tweak??

Post a reply to this message

Attachments:
Download 'gauss04e.jpg' (60 KB)

Preview of image 'gauss04e.jpg'

"Greg M. Johnson" wrote:
> 
> I also have an idea about using a "radiosity light bulb"-- perhaps not
> an original one.   When I tried it, I got just a few big white spots and
> then only after putting ambient to like rgb 200 or something.   Has
> anyone ever had success here;  any suggestions on which radiosity
> parameters to tweak??

Count. The smaller your sources of "light" are, the higher count you
need to use to ensure that the surfaces you're illuminating don't have
their random radiosity-testing rays spinning off into empty space, and a
lightbulb is very small indeed.

In fact, I would even suggest that you NOT attempt this. When I tried
it, I found that even increasing count to the max didn't provide results
that were acceptably smooth. Instead, use a normal light (perhaps with
fading) for your lightbulb, and then apply radiosity to that scene..
that way, a much larger part of the scene will already be lit, and
radiosity will be far more effective.

-Xplo

Post a reply to this message

Greg M. Johnson <"gregj;-)56590\""@aol.c;-)om> wrote in message
news:3b4768d1@news.povray.org...
> Here's some experimentation with use of radiosity to imply a "light
> source over there". It uses a large inverted sphere with a gradient
> pigment that is black to the left and white to the right.

I did a similar thing using a sky_sphere. What I found is that the brightness of
the object has to be adjusted relative to it's size, so a smaller lightsource
needs to be brighter to have the same effect on the scene (obvious if you think
about it!). Though I prefer using a light_source, as that way you can have
specular highlights and photons.

--
Tek
http://www.evilsuperbrain.com

Post a reply to this message

here is a scene I did. No light. The trick to have this "shadows" is to 
put a very high ambient object to cast the "light"...here, the window is 
a white box with ambient set to 100.
It's the technic I use to produce radiosity lights...It's working quite 
good even if you need to do a very high count. To get the effect you 
want, you can add a real light with a small intensity (rgb 0.25 or less) 
to help the radiosity process and get a better result with less render time.

      David

Post a reply to this message

Attachments:
Download 'appart.jpg' (23 KB)

Preview of image 'appart.jpg'

I'd have to say that the white on the floor and table looks especially odd.
Instead of being white, they would just be more illuminated.

Post a reply to this message

No actually on a bright day that's quite accurate, perhaps if the "light
box" he used had a slightly yellowish tint it would look more like sun,
otherwise it looks good to me.

"Ross Litscher" <lit### [at] osuedu> wrote in message
news:3b4a0e23@news.povray.org...
> I'd have to say that the white on the floor and table looks especially
odd.
> Instead of being white, they would just be more illuminated.
>
>

Post a reply to this message

"Greg M. Johnson" wrote:
> 
> Here's some experimentation with use of radiosity to imply a "light
> source over there". It uses a large inverted sphere with a gradient
> pigment that is black to the left and white to the right.
> 
> I also have an idea about using a "radiosity light bulb"-- perhaps not
> an original one.   When I tried it, I got just a few big white spots and
> then only after putting ambient to like rgb 200 or something.   Has
> anyone ever had success here;  any suggestions on which radiosity
> parameters to tweak??
> 

Apart from what others suggested, you should also reduce 'adc_bailout',
furthermore you could try to use negative ambient for the dark side, i
never tried that but it could work.

Christoph

-- 
Christoph Hormann <chr### [at] gmxde>
IsoWood include, radiosity tutorial, TransSkin and other 
things on: http://www.schunter.etc.tu-bs.de/~chris/

Post a reply to this message

For the table maybe, depending on the finish. But a carpet? It all still
looks odd to me :)

Thomas Lake <tla### [at] homecom> wrote in message
news:3b4a3e86$1@news.povray.org...
> No actually on a bright day that's quite accurate, perhaps if the "light
> box" he used had a slightly yellowish tint it would look more like sun,
> otherwise it looks good to me.
>
> "Ross Litscher" <lit### [at] osuedu> wrote in message
> news:3b4a0e23@news.povray.org...
> > I'd have to say that the white on the floor and table looks especially
> odd.
> > Instead of being white, they would just be more illuminated.
> >
> >
>
>

Post a reply to this message

Ross Litscher <lit### [at] osuedu> wrote in message
news:3b4cb37d$1@news.povray.org...
> For the table maybe, depending on the finish. But a carpet? It all still
> looks odd to me :)

It's a polished wooden floor, not a carpet. So what you see is a reflection
rather than illumination by the light. To make it look more like a real-world
highlight it should be slightly blurred, at present it looks like everything's
unrealistically perfectly smooth.

--
Tek
http://www.evilsuperbrain.com

Post a reply to this message