Hello
To study different designs and color combinations I tried to model a machine
we
are developing in my company. But I'm struggling a bit with the lighting and
look of materials.

Example picture: http:\\home.tiscalinet.ch\homes\christoph.boos\pov\lam2.png
pov code: http:\\home.tiscalinet.ch\homes\christoph.boos\pov\lam2.pov

I use kind of the same ideas for radiosity, floor and sky-sphere background
as
it was posted in aug 2002 by Jonathan "JRG" in a thread called "lightsetting
for
Lego Ferrari". I want to keep the focus on the machine and therefore
everything
else a simple white .(no fancy marble floor with cloudy sky background). The
ferrari example does it fine and I have to say I like the radiosity lighting
very much. But there are still some things I'm not completely happy with.

Right now my idea is to "dress" the machine with two materials. The main
part is
painted metal sheet casing (#declare design1) and then there is a cover made
of
"natural" (colorless) anodized aluminium. (#declare design2).

Now my questions:
- where does the light come from? The ambient value in the global_settings
as
well as in the material and floor settings are set to 0, I don't use any
other
light source...what is left  to emit light?

- Right now the most light seems to come from the top. There the machine is
very
 white but on the side it's more grayis, even the material is defined as
pure, clean, inocent white. Of course that's the point of raytracing to have
shadows and therefore brighter and darker parts but I would like have to
have
the main face (the large side with the aluminium cover) brighter and maybe
the
top a bit darker. I tried to add a light source but imidiately everything
got
very bright and lost contrast. (whitewhite machine on a whitewhite ground)

- To make the painted metal sheet a bit less artificial and give it a touch
of
the structure of the painting, I wanted to make this surface a bit rough.
But it
seems neither the roughness nor the crand seem to have an influence with
just
radiosity light. Other possibity to get a realistic painted and structured
surface?


Best regards
Christoph

Post a reply to this message

"Christoph" <chr### [at] switzerlandorg> wrote in message
news:web.3e346efac89d6394e499f8e40@news.povray.org...
>
> Example picture:
http:\\home.tiscalinet.ch\homes\christoph.boos\pov\lam2.png
> pov code: http:\\home.tiscalinet.ch\homes\christoph.boos\pov\lam2.pov

I can't get to those, only can get to the originating site and nothing
deeper than that.

> - where does the light come from?

'diffuse' in 'finish' is the major controlling factor.

> seems neither the roughness nor the crand seem to have an influence

Make sure the 'normal' keyword is in the radiosity statement so texture
normals are taken into account.

Would be nice to see those files.

Post a reply to this message

hughes, b. wrote:
>"Christoph" <chr### [at] switzerlandorg> wrote in message
>news:web.3e346efac89d6394e499f8e40[at]news.povray.org...
>>
>> Example picture:
>http:\\home.tiscalinet.ch\homes\christoph.boos\pov\lam2.png
>> pov code: http:\\home.tiscalinet.ch\homes\christoph.boos\pov\lam2.pov
>
>I can't get to those, only can get to the originating site and nothing
>deeper than that.

Sorrrrry, I got confused with the path... One more time:
Picture: http://home.tiscalinet.ch/christoph.boos/pov/lam2.png
File: http://home.tiscalinet.ch/christoph.boos/pov/lam2.pov

>> - where does the light come from?
>
>'diffuse' in 'finish' is the major controlling factor.

I see. Does that mean elements with a 'diffuse' in the finish emit light?

>> seems neither the roughness nor the crand seem to have an influence
>
>Make sure the 'normal' keyword is in the radiosity statement so texture
>normals are taken into account.

I'll try this. Just 'normal' without any parameter ? (I don't have the
manual next to me right now...)

Thanks
Christoph

Post a reply to this message

"Christoph" <chr### [at] switzerlandorg> wrote in message
news:web.3e34e4047108bb888c01fc670@news.povray.org...

> Picture: http://home.tiscalinet.ch/christoph.boos/pov/lam2.png
> File: http://home.tiscalinet.ch/christoph.boos/pov/lam2.pov
>
> Does that mean elements with a 'diffuse' in the finish emit light?

I'd say that, yes. Think of diffuse as the ability to reflect pigments, I
guess. Radiosity acts on the diffuse properties of the texture in a way so
that the color shines, similar to what you see from brightly colored objects
in sunlight when a darker room is surrounding them. More precisely, I guess
you could say it's a reflective thing, not a illumination thing, and
therefore indirect illumination.

> Just 'normal' without any parameter ?

radiosity {
RADIOSITY ITEMS
normal on
}

Sorry I didn't say that before, most such keywords use on/off (boolean)
although not all need it. I believe this one needs it. However, now having
looked at your scene file, I didn't find any normals in the textures.
 :-)

If I were better with radiosity myself I could be of more help. I can at
least suggest you might try recursion_limit 2 or 3, not just the 1 you have
in it now. And maybe max_samples could be of some use too,

I got a decent render by lowering diffuse to 0.7 and with the setup below, I
just am not sure what it should look like exactly. Only that your rendering
had a oversaturation.

global_settings {
 assumed_gamma 1
 ambient_light 0
 max_trace_level 15
 adc_bailout 1/150
 radiosity {
  brightness 1.2
  pretrace_start 0.08
  pretrace_end 0.01
  count 150 //
  error_bound 0.1
  recursion_limit 3
  normal on
  max_sample 0.8
 }
}

Post a reply to this message

Arg! I left out the fact that I had also changed the sky_sphere to:

sky_sphere {
 pigment {
  gradient y
  color_map {
   [0 color  rgb .75]
   [.5 color  rgb 1]
   [1 color  rgb .75]
  }
 }
}

And added a normal to the elox texture:

 normal {bumps 0.05 scale 0.01} // just to have one present

And.... your reflection with falloff 4 didn't make sense to me without a
variation from min to max:

 reflection {0.01, 0.1  falloff 4}

I think I'm done now  :-)

Post a reply to this message

hughes, b. wrote:
>Arg! I left out the fact that I had also changed the sky_sphere to:
>
>sky_sphere {
> ....

Aha... I think I get the idea. This could help...

>I got a decent render by lowering diffuse to 0.7 and with the setup below, I
>just am not sure what it should look like exactly.

Well, how should it look like...hard to describe.
It should look like a white machine on white floor with white background.
Now you may say that's like drawing with a white colored pencil on white
paper. But in reality the light intensity is always a bit different from
every side i.e. you always see shadows and different tints in a perspective
view. And I think here is the difficult thing. How to give it the look of a
white (and not gray) machine but on the same time still be able to see
edges, shapes and contrast to the background.
If you like look at this (real) foto, that's what I'm looking for:
http://home.tiscalinet.ch/christoph.boos/pov/studer.gif


>Only that your rendering had a oversaturation.
That means?


Thanks for your help
Christoph

Post a reply to this message

"Christoph" <chr### [at] switzerlandorg> wrote in message
news:web.3e3525e47108bb888c01fc670@news.povray.org...
> It should look like a white machine on white floor with white background.
> If you like look at this (real) foto, that's what I'm looking for:
> http://home.tiscalinet.ch/christoph.boos/pov/studer.gif
>
> > Only that your rendering had a oversaturation.
> That means?

That the top part of the machine was totally white, as a photo being
overexposed. Now, having seen that reference picture, it is just about the
same with a completely white top part on the right side. Only that the rest
of it is both plenty bright and dark both.

It isn't what I was starting to expect, based on your description just now
of what you wanted to do. I was thinking you were trying to get a very light
pencil drawing.

What you might have to try is different skies, or place a bright white
sphere or box out of sight of the camera but nearby enough to fake a light
source. In that picture you can see shadows caused by a soft light being
over on the left somewhere.

If you have trouble with the ground plane + sky sphere meeting on the
horizon (that the camera sees) you could use a very large sphere, and clip
it with a plane, for the ground instead. Setting it so the bottom-inside is
where the floor should be.

Post a reply to this message

>Well, how should it look like...hard to describe.
>It should look like a white machine on white floor with white background.
>Now you may say that's like drawing with a white colored pencil on white
>paper. But in reality the light intensity is always a bit different from
>every side i.e. you always see shadows and different tints in a perspective
>view. And I think here is the difficult thing. How to give it the look of a
>white (and not gray) machine but on the same time still be able to see
>edges, shapes and contrast to the background.
>If you like look at this (real) foto, that's what I'm looking for:
>http://home.tiscalinet.ch/christoph.boos/pov/studer.gif


Keep in mind, that "photograph" has probably had the machine pasted into the
white background in a paint program, particularly since the reflective
parts show darker shapes.  So try putting some other colors or textures
around the perimeter, out of view of the camera.  Radiosity will react to
it even if it's not visible in the scene.

In real life, I've never seen a completely white room.

Post a reply to this message