Hi,
I guess it somehow optimizes the volumetric sampling. I have no idea
what's going on 'under the hood'. Vesa..?
The idea is actually simple. Fog materials often require randomization of
the sampled position - this reduces banding. Randimization is simply done
as:
sample_position = original_position + random_delta
Purely position related effects, such as shadows and illumination in
general, automatically take the new position correctly into account.
However, some effects may be defined using 'Map coords' channel which is
computed before the randomization happens. In other words, the rendering
pipeline proceeds as:
Ray tracer: render engine computes position
Material mapping: parallel map computes Map coords from position
Material:
Turbidity is computed from Map coords
Position changes by randomization
Ray tracer: render enginer computes lighting using randomized
position
In above, turbidity and lighting are not computed using the same position.
Often a small difference does not matter, because the effect is quite
blurry. However, a perfectly exact method is this:
Ray tracer: render engine computes position
Material mapping: parallel map computes Map coords from position
Material:
Position changes by randomization
Map VSL object recomputes Map coords from random
position
Turbidity is computed from the new Map coords value
Ray tracer: render enginer computes lighting using randomized
position
I hope this makes the purpose of the Map object clear.
Best regards,
Vesa
