Hello everybody.
I'm trying to make the effect of an object burning and falling from the sky
so it should have a smoke tail (I guess you can imaging what Im talking
about). I thought I great idea would be to make a mesh with Wings3D and
then do something like:

interior{media{absorption 3}}

probably with density as well. But IT DOESN'T WORK. An here it goes my first
question DO MESH SUPPORT MEDIA?.

The second thing is that I know emission and scattering depend strongly on
the position of both light and camera. Absorption depends as well on the
relative position of light and camera?

Thank you

Post a reply to this message

kike <dry### [at] hotmailcom> wrote:
> interior{media{absorption 3}}

> probably with density as well. But IT DOESN'T WORK. An here it goes my first
> question DO MESH SUPPORT MEDIA?.

  Did you test that the exact same interior definition works ok with
some other object (such as a sphere)? You should make sure that the
problem is not in your media definition before making the conclusion
that the problem is with the mesh.

  If the mesh forms a closed surface then, AFAIK, media should work just
fine with it. (This is what I remember. I might have to actually try it
in practice in case I remember wrong.)

-- 
                                                          - Warp

Post a reply to this message

Thanks Warp
Yeah, you are completely right. I have just done what you recommend using a
sphere instead of a mesh, and the mesh doesnt seem to be the problem. It
should be the position of both camera and lights.

Post a reply to this message

web.45e6ccb2e5bdf9d1be7bfb550@news.povray.org...
>
> Thanks Warp
> Yeah, you are completely right. I have just done what you recommend using 
> a
> sphere instead of a mesh, and the mesh doesnt seem to be the problem. It
> should be the position of both camera and lights.
>
Are you sure you've put the keyword 'hollow' in your object definition?
Media can only be computed in a hollow object

Marc

Post a reply to this message

kike nous apporta ses lumieres en ce 01-03-2007 05:14:
> Hello everybody.
> I'm trying to make the effect of an object burning and falling from the sky
> so it should have a smoke tail (I guess you can imaging what Im talking
> about). I thought I great idea would be to make a mesh with Wings3D and
> then do something like:

> interior{media{absorption 3}}

> probably with density as well. But IT DOESN'T WORK. An here it goes my first
> question DO MESH SUPPORT MEDIA?.

> The second thing is that I know emission and scattering depend strongly on
> the position of both light and camera. Absorption depends as well on the
> relative position of light and camera?

> Thank you


You need to add hollow to the object's deffinition.
If you use the spherical pattern, remember that it's centered at the origin, 
value of 1 at <0,0,0>, droping to 0 at a radius of 1 unit.
If you use the cylindrical pattern, it's centered around the y axis with an 
effective unit radius.
It's beter to create your object at that location and orientation, then to 
rotate and translate to the desired orientation and location.

Emissive media don't interact with your light. It's aspect depent on it's 
colour, it's density and thickness. It don't cast any shadow. Is visible on it's 
own.

Absorbing media will absorb light, will cast shadow and obscure what is behind. 
Aspect is independant on the direction of incident light. It needs something 
behind it to be obscured to be visible.

Scattering media will diffuse light and obscure what is behind it. The 
directionality depends on the scatering model used. Type 1 is isotropic or 
spherical scatering. It's mostly independent of the light's direction relative 
to the camera, it will show some darkening gradient if thick enough in the 
direction of the incoming light.
Type 2 and 3 are strongly directionals. Low effect if the light is behind the 
camera, maximum if you look at the light.
Type 4 have some directionality. Maximum effec if the light is behind or 
directly in front of the camera, less but notable if perpendicular.
Type 5's directionality depends on the exentricity parameter.
Those only affect scatering media, NOT emissive nor absorbing media.
Light shining through it will make it visible as well as any background as it is 
also absorbing.

-- 
Alain
-------------------------------------------------
"If you see me running, try to keep up."
       ...Back of bomb technician's shirt

Post a reply to this message

Thanks to all of you. I used hollow from the very begining (I'm new user but
not so new) but that was not the problem. Finally Alain was right, the
problem was the background.

"Absorbing media will absorb light, will cast shadow and obscure what is
behind.
 Aspect is independant on the direction of incident light. It needs
something
 behind it to be obscured to be visible."

But sadly, background can not be changed in my image, so I will have to
change the way I model the smoke. Any idea? Anyone knows how smoke was done
in the image of Pearl Harbour?
http://hof.povray.org/7b.html
It seems to be done with small spheres...

---------------------------
Kike
My raytracing jobs are in:
http://personales.ya.com/robertoullan/

Post a reply to this message

kike <dry### [at] hotmailcom> wrote:
> But sadly, background can not be changed in my image, so I will have to
> change the way I model the smoke. Any idea?

  If your absorbing media is not showing up because of the background
color, it means that your background is black or very dark?

  If you think about it, are you even *supposed* to see (black) smoke
on a black background?-)

  If what you want is for light sources to "illuminate" your smoke,
then that's also possible, but you can't do it with absorbing media.
You need to use scattering media, and expect having to fine-tune it
a lot and a considerable increase in rendering times.

  If that's not what you want, but you still want your smoke to be
visible, then you might try to make the smoke slightly emissive instead.
In other words, besides being absorbing, make it also slightly emissive
(this is possible by putting *two* media blocks in the interior block,
one being absorbing and the other emissive). It might not get the effect
you want, but you might want to try it anyways.

-- 
                                                          - Warp

Post a reply to this message

No, my background is quite bright, nearly white. But when I said the problem
was the background I wanted to say not only the colour of the background but
the position of lights and camera. The image is supposed to be a dawn and
the camera is looking to the sun so the position of elements is

CAMERA-LIGHT-BACKGROUND(sky and clouds)

And I guess it is important where I should put my absorption media, I guess
it is not the same to put it between camera and light of between light and
background. Or at least thats what it seems to be.

In the first case the absorption media should avoid some rays to reach the
camera so you should see the effect of absorption. But I'm not completely
sure.

If I dont solve this problem this weekend I will put here the code to show
you exactly what Im talking about.

THANKS A LOT

Post a reply to this message

kike <dry### [at] hotmailcom> wrote:
> If I dont solve this problem this weekend I will put here the code to show
> you exactly what Im talking about.

  Why not create a small example scene with for example a sphere with
the absorption media you are using and a background colored like in your
scene and post it here? It would be easier to know what is the problem
you are having if you cold see it for ourselves.

-- 
                                                          - Warp

Post a reply to this message

I finally solved the problem. It was (as I suspected) a problem in between
relative position of light, camera and absorbing object. But now I have
another problem. Mesh supports absorption but I thought I could simulate
smoke with a mesh and now I see it is not possible. Let me explain this.

If you make an absorbing mesh, absorption will fix to the mesh but you will
not see the cool effect of noisy smoke in the edges. To see this cool
effect you have to define a density pattern (for example a spherical
density pattern) smaller than the mesh with some turbulence. But then what
you actually see is that the absorption has the same shape for a mesh, a
box or a sphere. So doesnt seem to be a good idea to sculpt a mesh with the
shape of a smoke column and then expect absorption to do the rest. It is
better to use the trick of using several spheres  (as the smoke plug-in I
have seen).

Post a reply to this message