[Bf-blender-cvs] [5f97b51] soc-2016-cycles_denoising: Cycles: Refactor storing of the integration result
Lukas Stockner
noreply at git.blender.org
Sat Jun 4 22:03:15 CEST 2016
Commit: 5f97b51d15cd9e5a1ddaa244ba73dba3d89beb16
Author: Lukas Stockner
Date: Sat Jun 4 21:50:11 2016 +0200
Branches: soc-2016-cycles_denoising
https://developer.blender.org/rB5f97b51d15cd9e5a1ddaa244ba73dba3d89beb16
Cycles: Refactor storing of the integration result
This commit refactors how the integration result is stored: Instead of summing up
the PathRadiance in the integration function and returning the final color, the integration
function now fills a PathRadiance passed to it and just returns the alpha value.
The main kernel function then passes that PathRadiance to kernel_write_result, a new function
which then handles summing, clamping and storing of light and combined passes.
This commit by itself shouldn't change existing behaviour, but is needed for the upcoming
selective denoising.
===================================================================
M intern/cycles/kernel/kernel_passes.h
M intern/cycles/kernel/kernel_path.h
M intern/cycles/kernel/kernel_path_branched.h
M intern/cycles/kernel/kernel_types.h
===================================================================
diff --git a/intern/cycles/kernel/kernel_passes.h b/intern/cycles/kernel/kernel_passes.h
index 5b65bff..32f0d68 100644
--- a/intern/cycles/kernel/kernel_passes.h
+++ b/intern/cycles/kernel/kernel_passes.h
@@ -295,5 +295,19 @@ ccl_device_inline void kernel_write_light_passes(KernelGlobals *kg, ccl_global f
#endif
}
+ccl_device_inline void kernel_write_result(KernelGlobals *kg, ccl_global float *buffer,
+ int sample, PathRadiance *L, float alpha)
+{
+ if(L) {
+ float3 L_sum = path_radiance_clamp_and_sum(kg, L);
+ kernel_write_pass_float4(buffer, sample, make_float4(L_sum.x, L_sum.y, L_sum.z, alpha));
+
+ kernel_write_light_passes(kg, buffer, L, sample);
+ }
+ else {
+ kernel_write_pass_float4(buffer, sample, make_float4(0.0f, 0.0f, 0.0f, 0.0f));
+ }
+}
+
CCL_NAMESPACE_END
diff --git a/intern/cycles/kernel/kernel_path.h b/intern/cycles/kernel/kernel_path.h
index 5c4fa42..3e8f9e8 100644
--- a/intern/cycles/kernel/kernel_path.h
+++ b/intern/cycles/kernel/kernel_path.h
@@ -591,18 +591,18 @@ ccl_device void kernel_path_subsurface_setup_indirect(
#endif /* __SUBSURFACE__ */
-ccl_device_inline float4 kernel_path_integrate(KernelGlobals *kg,
+ccl_device_inline float kernel_path_integrate(KernelGlobals *kg,
RNG *rng,
int sample,
Ray ray,
- ccl_global float *buffer)
+ ccl_global float *buffer,
+ PathRadiance *L)
{
/* initialize */
- PathRadiance L;
float3 throughput = make_float3(1.0f, 1.0f, 1.0f);
float L_transparent = 0.0f;
- path_radiance_init(&L, kernel_data.film.use_light_pass);
+ path_radiance_init(L, kernel_data.film.use_light_pass);
PathState state;
path_state_init(kg, &state, rng, sample, &ray);
@@ -670,7 +670,7 @@ ccl_device_inline float4 kernel_path_integrate(KernelGlobals *kg,
float3 emission;
if(indirect_lamp_emission(kg, &state, &light_ray, &emission))
- path_radiance_accum_emission(&L, throughput, emission, state.bounce);
+ path_radiance_accum_emission(L, throughput, emission, state.bounce);
}
#endif
@@ -699,7 +699,7 @@ ccl_device_inline float4 kernel_path_integrate(KernelGlobals *kg,
/* emission */
if(volume_segment.closure_flag & SD_EMISSION)
- path_radiance_accum_emission(&L, throughput, volume_segment.accum_emission, state.bounce);
+ path_radiance_accum_emission(L, throughput, volume_segment.accum_emission, state.bounce);
/* scattering */
VolumeIntegrateResult result = VOLUME_PATH_ATTENUATED;
@@ -709,7 +709,7 @@ ccl_device_inline float4 kernel_path_integrate(KernelGlobals *kg,
/* direct light sampling */
kernel_branched_path_volume_connect_light(kg, rng, &volume_sd,
- throughput, &state, &L, all, &volume_ray, &volume_segment);
+ throughput, &state, L, all, &volume_ray, &volume_segment);
/* indirect sample. if we use distance sampling and take just
* one sample for direct and indirect light, we could share
@@ -726,7 +726,7 @@ ccl_device_inline float4 kernel_path_integrate(KernelGlobals *kg,
kernel_volume_decoupled_free(kg, &volume_segment);
if(result == VOLUME_PATH_SCATTERED) {
- if(kernel_path_volume_bounce(kg, rng, &volume_sd, &throughput, &state, &L, &ray))
+ if(kernel_path_volume_bounce(kg, rng, &volume_sd, &throughput, &state, L, &ray))
continue;
else
break;
@@ -741,15 +741,15 @@ ccl_device_inline float4 kernel_path_integrate(KernelGlobals *kg,
/* integrate along volume segment with distance sampling */
ShaderData volume_sd;
VolumeIntegrateResult result = kernel_volume_integrate(
- kg, &state, &volume_sd, &volume_ray, &L, &throughput, rng, heterogeneous);
+ kg, &state, &volume_sd, &volume_ray, L, &throughput, rng, heterogeneous);
# ifdef __VOLUME_SCATTER__
if(result == VOLUME_PATH_SCATTERED) {
/* direct lighting */
- kernel_path_volume_connect_light(kg, rng, &volume_sd, throughput, &state, &L);
+ kernel_path_volume_connect_light(kg, rng, &volume_sd, throughput, &state, L);
/* indirect light bounce */
- if(kernel_path_volume_bounce(kg, rng, &volume_sd, &throughput, &state, &L, &ray))
+ if(kernel_path_volume_bounce(kg, rng, &volume_sd, &throughput, &state, L, &ray))
continue;
else
break;
@@ -811,7 +811,7 @@ ccl_device_inline float4 kernel_path_integrate(KernelGlobals *kg,
#endif
/* holdout mask objects do not write data passes */
- kernel_write_data_passes(kg, buffer, &L, &sd, sample, &state, throughput);
+ kernel_write_data_passes(kg, buffer, L, &sd, sample, &state, throughput);
/* blurring of bsdf after bounces, for rays that have a small likelihood
* of following this particular path (diffuse, rough glossy) */
@@ -829,7 +829,7 @@ ccl_device_inline float4 kernel_path_integrate(KernelGlobals *kg,
if(sd.flag & SD_EMISSION) {
/* todo: is isect.t wrong here for transparent surfaces? */
float3 emission = indirect_primitive_emission(kg, &sd, isect.t, state.flag, state.ray_pdf);
- path_radiance_accum_emission(&L, throughput, emission, state.bounce);
+ path_radiance_accum_emission(L, throughput, emission, state.bounce);
}
#endif
@@ -853,7 +853,7 @@ ccl_device_inline float4 kernel_path_integrate(KernelGlobals *kg,
#ifdef __AO__
/* ambient occlusion */
if(kernel_data.integrator.use_ambient_occlusion || (sd.flag & SD_AO)) {
- kernel_path_ao(kg, &sd, &L, &state, rng, throughput);
+ kernel_path_ao(kg, &sd, L, &state, rng, throughput);
}
#endif
@@ -863,7 +863,7 @@ ccl_device_inline float4 kernel_path_integrate(KernelGlobals *kg,
if(sd.flag & SD_BSSRDF) {
if(kernel_path_subsurface_scatter(kg,
&sd,
- &L,
+ L,
&state,
rng,
&ray,
@@ -876,15 +876,15 @@ ccl_device_inline float4 kernel_path_integrate(KernelGlobals *kg,
#endif /* __SUBSURFACE__ */
/* direct lighting */
- kernel_path_surface_connect_light(kg, rng, &sd, throughput, &state, &L);
+ kernel_path_surface_connect_light(kg, rng, &sd, throughput, &state, L);
/* compute direct lighting and next bounce */
- if(!kernel_path_surface_bounce(kg, rng, &sd, &throughput, &state, &L, &ray))
+ if(!kernel_path_surface_bounce(kg, rng, &sd, &throughput, &state, L, &ray))
break;
}
#ifdef __SUBSURFACE__
- kernel_path_subsurface_accum_indirect(&ss_indirect, &L);
+ kernel_path_subsurface_accum_indirect(&ss_indirect, L);
/* Trace indirect subsurface rays by restarting the loop. this uses less
* stack memory than invoking kernel_path_indirect.
@@ -894,7 +894,7 @@ ccl_device_inline float4 kernel_path_integrate(KernelGlobals *kg,
&ss_indirect,
&state,
&ray,
- &L,
+ L,
&throughput);
}
else {
@@ -903,15 +903,11 @@ ccl_device_inline float4 kernel_path_integrate(KernelGlobals *kg,
}
#endif /* __SUBSURFACE__ */
- float3 L_sum = path_radiance_clamp_and_sum(kg, &L);
-
- kernel_write_light_passes(kg, buffer, &L, sample);
-
#ifdef __KERNEL_DEBUG__
kernel_write_debug_passes(kg, buffer, &state, &debug_data, sample);
#endif
- return make_float4(L_sum.x, L_sum.y, L_sum.z, 1.0f - L_transparent);
+ return 1.0f - L_transparent;
}
ccl_device void kernel_path_trace(KernelGlobals *kg,
@@ -932,15 +928,15 @@ ccl_device void kernel_path_trace(KernelGlobals *kg,
kernel_path_trace_setup(kg, rng_state, sample, x, y, &rng, &ray);
/* integrate */
- float4 L;
-
- if(ray.t != 0.0f)
- L = kernel_path_integrate(kg, &rng, sample, ray, buffer);
- else
- L = make_float4(0.0f, 0.0f, 0.0f, 0.0f);
+ PathRadiance L;
- /* accumulate result in output buffer */
- kernel_write_pass_float4(buffer, sample, L);
+ if(ray.t != 0.0f) {
+ float alpha = kernel_path_integrate(kg, &rng, sample, ray, buffer, &L);
+ kernel_write_result(kg, buffer, sample, &L, alpha);
+ }
+ else {
+ kernel_write_result(kg, buffer, sample, NULL, 0.0f);
+ }
path_rng_end(kg, rng_state, rng);
}
diff --git a/intern/cycles/kernel/kernel_path_branched.h b/intern/cycles/kernel/kernel_path_branched.h
index a7953a0..850b3a0 100644
--- a/intern/cycles/kernel/kernel_path_branched.h
+++ b/intern/cycles/kernel/kernel_path_branched.h
@@ -222,14 +222,13 @@ ccl_device void kernel_branched_path_subsurface_scatter(KernelGlobals *kg,
}
#endif
-ccl_device float4 kernel_branched_path_integrate(KernelGlobals *kg, RNG *rng, int sample, Ray ray, ccl_global float *buffer)
+ccl_device float kernel_branched_path_integrate(KernelGlobals *kg, RNG *rng, int sample, Ray ray, ccl_global float *buffer, PathRadiance *L)
{
/* initialize */
- PathRadiance L;
float3 throughput = make_float3(1.0f, 1.0f, 1.0f);
float L_transparent = 0.0f;
- path_radiance_init(&L, kernel_data.film.use_light_pass);
+ path_radiance_init(L, kernel_data.film.use_light_pass);
PathState state;
path_state_init(kg, &state, rng, sample, &ray);
@@ -300,7 +299,7 @@ ccl_device float4 kernel_branched_path_integrate(KernelGlobals *kg, RNG *rng, in
int all = kernel_data.integrator.sample_all_lights_direct;
kernel_branched_path_volume_connect_light(kg, rng, &volume_sd,
- throughput, &state, &L, all, &volume_ray, &volume_segment);
+ throughput, &state, L, all, &volume_ray, &volume_segment);
/* indirect light sampling */
int num_samples = kernel_data.integrator.volume_samples;
@@ -336,7 +335,7 @@ ccl_device float4 kernel_branched_path_integrate(KernelGlobals *kg, RNG *rng, in
&volume_sd,
&tp,
&ps,
- &L,
+ L,
@@ Diff output truncated at 10240 characters. @@
More information about the Bf-blender-cvs
mailing list