Could someone explain me how I should read the scattering diagrams
in the "media" section of the POV-Ray documentation ?

I'd like to know, especially, what are the values on the axes, 
and how I should understand them.

Links to clear ressources about the various scattering functions
would also be appreciated.

Thanks,
Fabien.

Post a reply to this message

Fabien Mosen wrote:
> 
> Could someone explain me how I should read the scattering diagrams
> in the "media" section of the POV-Ray documentation ?
> 
> I'd like to know, especially, what are the values on the axes,
> and how I should understand them.

I think they are fairly easy to understand.  As the text next to the
diagram on the left says the light is coming from the left.  The curve
shows the intensity of scattered light depending of the camera.  If you
look staight against the light source the intensity is maximum in most
scattering models.  

The scale of the axis is simply fractions of the maximum intensity.  

The values on the horizontal axis mean:
left: intensity when light source is behind the camera
right: intensity when camera is pointing directly towards the light source

The values on the vertical axis mean:
camera viewing direction orthogonal to the illumination direction

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

Christoph Hormann wrote:

> Fabien Mosen wrote:
> >
> > Could someone explain me how I should read the scattering diagrams
> > in the "media" section of the POV-Ray documentation ?
> >
> > I'd like to know, especially, what are the values on the axes,
> > and how I should understand them.
>
> The values on the horizontal axis mean:
> left: intensity when light source is behind the camera
> right: intensity when camera is pointing directly towards the light source
>
> The values on the vertical axis mean:
> camera viewing direction orthogonal to the illumination direction

I don't know if this is correct.

The POV help file does not specify the axes, and they appear to be cartesian.
However, I worked with these diagrams during the summer, and they were always
in polar coordinates.

If this applies to the POV diagrams (and from their shape, I think it does),
then the distance of the point from the (0,0) point is an indication of
intensity, and the angle is the observation angle.

For the Rayleigh Scattering, for example, the diagram means that most of the
light is scattered forwards and backwards (in equal amounts, since the plot is
symmetric), with less scattered perpendicular to the direction of light.

The Mie Scattering, on the other hand is not symmetrical, and scatters more
favourably in the forward direction than back towards the light source. It
also scatters perpendicularly, but the narrowness of the curve shows that this
is much less than for the Rayleigh Scattering.

If one were to use the Heyney-Greenstein model with an eccentricity of 0 (a
perfect circle), then all light would be scattered equally in all directions.

Keep in mind that these diagrams are in two dimensions, while real scattering
takes place in three dimensions. I'm not sure of the diagrams are already
corrected for the integration required for this translation (since the light %
would be different if the dimensions are increased, I think) but they should
be roughly the same anyways.

I hope this is of some help.


Simon

Post a reply to this message

Simon de Vet wrote:

> The POV help file does not specify the axes, and they appear to be cartesian.
> However, I worked with these diagrams during the summer, and they were always
> in polar coordinates.

I should clarify that I did not work with the POV diagrams (I never paid them any
attention before today), but I worked with similar scattering diagrams in a
university research setting.


Simon

Post a reply to this message

Simon de Vet wrote:
> 
> The POV help file does not specify the axes, and they appear to be cartesian.
> However, I worked with these diagrams during the summer, and they were always
> in polar coordinates.
> 
> If this applies to the POV diagrams (and from their shape, I think it does),
> then the distance of the point from the (0,0) point is an indication of
> intensity, and the angle is the observation angle.
> 

I understood it that way too, just like directional characteristics of
radio antennas.  

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