9.5.2

356 • CHAPTER 7
OBJECTS MENU • LIGHTING • 357
So why is volumetric lighting so time-consuming for the renderer?
When a beam hits a light cone, it is not only the intensity of the light that needs to be computed.
Additionally, for each part of the beam, CINEMA 4D needs to look for other objects within the light
cone that might be casting shadows. So for every part of the beam of light, an extra raytracer ray needs
to be initiated and emitted. Because it is not possible to shrink segments in the fog below a certain
length, an approximation must be used; the length of the light cone is subdivided into equal parts.
Suppose the raytracer ray hits the light cone and the distance between the entry and the exit points
of the light cone is 1,000 units. A sample distance of 50 units will mean that an intensity value and a
shadow beam will have to be calculated 20 times (1,000/50).
The shorter the sample distance, the longer the calculation will take. Even with just ve subdivisions
(so a sample distance of 200 in the above example), this will require a ve-fold increase per raytracer
ray and per contact with the light cone than without volumetric lighting. Using progressively ner
subdivisions, the processing time involved very quickly becomes astronomical.
This is an inherent problem with computer graphics that cannot be resolved or accelerated other than
by throwing processor power at it.
Volumetric lighting needs a lot of calculation time, therefore render such light
sources only when it is absolutely necessary.
So why can’t you input a xed value for the number of samples?
Well, if the raytracer beam hits the light cone at its beginning, the distance between the entry and exit
points might be, for example, 100 units. But if the beam hits the cone further from the light source,
this distance might grow to 5,000 units or more. So if you used a xed number of samples, at the
narrow end of the cone a lot of unnecessary calculations would be made and later too few (which
would result in artefacts).
Brightness
This value is the brightness of the visible light source.
Dust
Determines the darkness of the light cone. With a Dust value of more than 0%, Brightness is subtracted
instead of added. To ensure you see the full effect of this, lower the light’s brightness accordingly. The
difference between a normal bright light and a dust-assigned light can be seen clearly in the following
illustration. To the left is a bright, visible light. To the right a dark, dusty, somewhat sooty light.