Among other things, wrote:

> refracted light is not governed by fade_distance/fade_power is it? i think
> i recall reading that in the docs of 3.5. (opening docs for 3.6...) i
> always wondered why. for instance if you have a candle (point light) in a
> glass candle holder, nearly all the light would be defracted, and
> therefore, not fade with distance.
> 
> from the documentation "Reflected or refracted light is not attenuated by
> distance." is this because it is computationally too expensive, or is this
> just how light works in nature?

I will guess an explanation.

The "physical" fade_power=2 comes, as explained, from a geometric/topologic 
fact and, ultimately, from the light traveling in (almost) straight lines. 
If light travels in straight lines, in all directions, the "amount" of 
light per unit area (the perceived lighting intensity) diminishes with the 
square of distance, just because the surface (of a sphere, for example) 
grows with the square of distance (the sphere's radius).

This effect could be obtained by using forward raytracing: Shoot light rays 
in every direction, then count how many rays reach a given 
object/area/surface. Using photons in POV-Ray would, I guess, reproduce the 
"inverse square law" without need of any kind of fade_power, because it's 
intrinsic to the way light propagates (and it's correctly simulated in this 
case).

With backward raytracing, however, there is no easy way to know how many of 
the "original" light rays will have reached a given object, all you know is 
it is possible for a ray to reach the object. But, assuming the light 
travels in straight line, it is safe (and realistic) to just use a 
fade_power=2 to simulate the light fading. As soon as the light doesn't 
travel in a straight line, which happens when it is "reflacted" or 
"refrected" the assumption is not valid and the "fade_power=2" could be 
plain wrong, you have to take into account all the possible ways a ray 
could reach the object.

Summing it up: fade_power=2 is a quick and useful way to simulate light when 
it travels in straight lines, but it cannot be used when it doesn't!

Now, this doesn't mean that's the actual reason it works like that in 
POV-Ray. I haven't thought quite deeply about the physics and the "POV-way" 
in this phenomenon and I could be wrong.

-- 
light_source{9+9*x,1}camera{orthographic look_at(1-y)/4angle 30location
9/4-z*4}light_source{-9*z,1}union{box{.9-z.1+x clipped_by{plane{2+y-4*x
0}}}box{z-y-.1.1+z}box{-.1.1+x}box{.1z-.1}pigment{rgb<.8.2,1>}}//Jellby

Post a reply to this message