Quality Control - Photographic Rendering with V-Ray for SketchUp - page 307

The premise of the Sample Rate channel is very simple; it is designed to show us
where the available samples or rays (the number of which is controlled by the main
image sampling options) are being used in the scene.
Dark blue denotes areas where a minimum or low number of samples are being em-
ployed, light blue through to green shows where a higher number of samples are
being employed, while yellow to orange and then into red tell us that an increasingly
high and ultimately maximum number of samples are having to be used. This means
that image sampling is working very hard to try and resolve aliasing and noise is-
sues, and where areas of pure red exist, there may actually not be enough samples
available to adequately resolve the artifact problems.
The information in this image gives us the ability to fine tune the balance in the use
of global samples (controlled by the global Image sampler settings) as opposed to
the local or secondary samples coming from elements such as materials and lights
for example (we have already mentioned this a couple of times in this chapter). Any
areas where global sampling is being heavily employed (red pixels) are good candid-
ates for increasing the local samples so as to balance out the load and improve the
overall quality of the final image, often without noticeably increasing render times.
One instantly noticeable piece of information in the Sample Rate channel would be
its agreement with our earlier observation from the RGB image that the frosted glass
and metal frame of the door are quite noisy, since they were sampled quite substan-
tially (as denoted by the green and even red colored areas).
The image also shows that our floor and wall materials are also require quite a bit of
help from image sampling, which is perhaps not surprising given that they both have
blurry reflections applied. So, we may want to look at increasing local subdivs there.
Seeing red around the edges of the direct light hotspots and in the bright reflections
on the chrome is not really surprising as image sampling struggles to average out
extremely bright pixel values. So, this is not necessarily something to be overly con-