Image Space Gathering

Soft shadows, glossy reflections and depth of field are valuable effects
for realistic rendering and are often computed using distribution
ray tracing (DRT). These “blurry” effects often need not be
accurate and are sometimes simulated by blurring an image with
sharper effects, such as blurring hard shadows to simulate soft shadows.
One of the most effective examples of such a blurring algorithm
is percentage closer soft shadows (PCSS). That technique,
however, does not naturally extend to shadows generated in image
space, such as those computed by a ray tracer, nor does it extend to
glossy reflections or depth of field. This limitation can be overcome
by generalizing PCSS to be phrased in terms of a gather from image
space textures implemented with cross bilateral filtering. This
paper demonstrates a framework to create visually compelling and phenomenologically accurate approximations of DRT effects based
on repeatedly gathering from bilaterally weighted image space texture
samples. These gathering and filtering operations are well supported
by modern parallel architectures, enabling this technique to
run at interactive rates.