Blender Git Loki

Adaptive Sampling for Cycles.

This feature takes some inspiration from
"RenderMan: An Advanced Path Tracing Architecture for Movie Rendering" and
"A Hierarchical Automatic Stopping Condition for Monte Carlo Global Illumination"

In addition, it implements the sampling pattern described in
"Progressive Multi-Jittered Sample Sequences"

The basic principle is as follows:
While samples are being added to a pixel, the adaptive sampler writes half
of the samples to a separate buffer. This gives it two separate estimates
of the same pixel, and by comparing their difference it estimates convergence.
Once convergence drops below a given threshold, the pixel is considered done.

When a pixel has not converged yet and needs more samples than the minimum,
its immediate neighbors are also set to take more samples. This is done in order
to more reliably detect sharp features such as caustics. A 3x3 box filter that
is run peridoically over the tile buffer is used for that purpose.

After a tile is finished rendering, the values of all passes are scaled as if
they were rendered with the full number of samples. This way, any code operating
on these buffers, for example the denoiser, does not need to be changed for
per-pixel sample counts.

Differential Revision: https://developer.blender.org/D4686

Cycles: Moved "Adaptive Sampilng" option out of the advanced section in the UI.

Formatting: Eliminated some superfluous white space.

Reverted accidental change.

Merge branch 'master' into cycles_adaptive_sampling

Variable naming, white space fixes.

Cycles: Adaptive sampling auto paramters now print to the console.

Cycles: Fixed adaptive sampling with progressive refine on CUDA devices.

Cycles: Some refactoring and improved shuffle in PMJ02 sampler.

Cycles: PMJ sequence generation now happens in multiple threads.

Merge branch 'cycles_adaptive_sampling' of https://github.com/skwerner/blender into cycles_adaptive_sampling

Cycles: Renamed sobol_directions to sample_pattern_lut since PMJ is using it too.

Cycles: Renamed sobol_directions to sample_pattern_lut since PMJ is using it too.

Cycles: Fixed progressive rendeing with adaptive sampling.

Cycles: Adaptive sampling support for split kernel/OpenCL.

Cycles: Started work on split kernel support for adaptive sampling.

Cycles: Fixed adaptive sampling with CUDA, some work on adaptive sampling with progressive refine (not finished yet).

Cycles: Creating extra passes for adaptive sampling only when it's actually enabled.

Cycles: Fixed OpenCL kernel build.

Cycles: Improved adaptive sampling for CUDA, some more refactoring.