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"

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 periodically over the tile buffer is used for that purpose.

After a tile has 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.

Reviewed By: brecht, #cycles

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

4 files changed, 140 insertions, 0 deletions

diff --git a/intern/cycles/kernel/split/kernel_adaptive_adjust_samples.h b/intern/cycles/kernel/split/kernel_adaptive_adjust_samples.h
new file mode 100644
index 00000000000..60ebf415970
--- /dev/null
+++ b/intern/cycles/kernel/split/kernel_adaptive_adjust_samples.h
@@ -0,0 +1,44 @@
+/*
+ * Copyright 2019 Blender Foundation
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+CCL_NAMESPACE_BEGIN
+
+ccl_device void kernel_adaptive_adjust_samples(KernelGlobals *kg)
+{
+  int pixel_index = ccl_global_id(1) * ccl_global_size(0) + ccl_global_id(0);
+  if (pixel_index < kernel_split_params.tile.w * kernel_split_params.tile.h) {
+    int x = kernel_split_params.tile.x + pixel_index % kernel_split_params.tile.w;
+    int y = kernel_split_params.tile.y + pixel_index / kernel_split_params.tile.w;
+    int buffer_offset = (kernel_split_params.tile.offset + x +
+                         y * kernel_split_params.tile.stride) *
+                        kernel_data.film.pass_stride;
+    ccl_global float *buffer = kernel_split_params.tile.buffer + buffer_offset;
+    int sample = kernel_split_params.tile.start_sample + kernel_split_params.tile.num_samples;
+    if (buffer[kernel_data.film.pass_sample_count] < 0.0f) {
+      buffer[kernel_data.film.pass_sample_count] = -buffer[kernel_data.film.pass_sample_count];
+      float sample_multiplier = sample / max((float)kernel_split_params.tile.start_sample + 1.0f,
+                                             buffer[kernel_data.film.pass_sample_count]);
+      if (sample_multiplier != 1.0f) {
+        kernel_adaptive_post_adjust(kg, buffer, sample_multiplier);
+      }
+    }
+    else {
+      kernel_adaptive_post_adjust(kg, buffer, sample / (sample - 1.0f));
+    }
+  }
+}
+
+CCL_NAMESPACE_END
diff --git a/intern/cycles/kernel/split/kernel_adaptive_filter_x.h b/intern/cycles/kernel/split/kernel_adaptive_filter_x.h
new file mode 100644
index 00000000000..93f41f7ced4
--- /dev/null
+++ b/intern/cycles/kernel/split/kernel_adaptive_filter_x.h
@@ -0,0 +1,30 @@
+/*
+ * Copyright 2019 Blender Foundation
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+CCL_NAMESPACE_BEGIN
+
+ccl_device void kernel_adaptive_filter_x(KernelGlobals *kg)
+{
+  int pixel_index = ccl_global_id(1) * ccl_global_size(0) + ccl_global_id(0);
+  if (pixel_index < kernel_split_params.tile.h &&
+      kernel_split_params.tile.start_sample + kernel_split_params.tile.num_samples >=
+          kernel_data.integrator.adaptive_min_samples) {
+    int y = kernel_split_params.tile.y + pixel_index;
+    kernel_do_adaptive_filter_x(kg, y, &kernel_split_params.tile);
+  }
+}
+
+CCL_NAMESPACE_END
diff --git a/intern/cycles/kernel/split/kernel_adaptive_filter_y.h b/intern/cycles/kernel/split/kernel_adaptive_filter_y.h
new file mode 100644
index 00000000000..eca53d079ec
--- /dev/null
+++ b/intern/cycles/kernel/split/kernel_adaptive_filter_y.h
@@ -0,0 +1,29 @@
+/*
+ * Copyright 2019 Blender Foundation
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+CCL_NAMESPACE_BEGIN
+
+ccl_device void kernel_adaptive_filter_y(KernelGlobals *kg)
+{
+  int pixel_index = ccl_global_id(1) * ccl_global_size(0) + ccl_global_id(0);
+  if (pixel_index < kernel_split_params.tile.w &&
+      kernel_split_params.tile.start_sample + kernel_split_params.tile.num_samples >=
+          kernel_data.integrator.adaptive_min_samples) {
+    int x = kernel_split_params.tile.x + pixel_index;
+    kernel_do_adaptive_filter_y(kg, x, &kernel_split_params.tile);
+  }
+}
+CCL_NAMESPACE_END
diff --git a/intern/cycles/kernel/split/kernel_adaptive_stopping.h b/intern/cycles/kernel/split/kernel_adaptive_stopping.h
new file mode 100644
index 00000000000..c8eb1ebd705
--- /dev/null
+++ b/intern/cycles/kernel/split/kernel_adaptive_stopping.h
@@ -0,0 +1,37 @@
+/*
+ * Copyright 2019 Blender Foundation
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+CCL_NAMESPACE_BEGIN
+
+ccl_device void kernel_adaptive_stopping(KernelGlobals *kg)
+{
+  int pixel_index = ccl_global_id(1) * ccl_global_size(0) + ccl_global_id(0);
+  if (pixel_index < kernel_split_params.tile.w * kernel_split_params.tile.h &&
+      kernel_split_params.tile.start_sample + kernel_split_params.tile.num_samples >=
+          kernel_data.integrator.adaptive_min_samples) {
+    int x = kernel_split_params.tile.x + pixel_index % kernel_split_params.tile.w;
+    int y = kernel_split_params.tile.y + pixel_index / kernel_split_params.tile.w;
+    int buffer_offset = (kernel_split_params.tile.offset + x +
+                         y * kernel_split_params.tile.stride) *
+                        kernel_data.film.pass_stride;
+    ccl_global float *buffer = kernel_split_params.tile.buffer + buffer_offset;
+    kernel_do_adaptive_stopping(kg,
+                                buffer,
+                                kernel_split_params.tile.start_sample +
+                                    kernel_split_params.tile.num_samples - 1);
+  }
+}
+CCL_NAMESPACE_END