This caused a couple of fireflies in koro_final.blend. The wrong normal would
cause the shading point to be set as backfacing, which triggered another bug
with hair BSDFs on the backface of hair curves. That one is not fixed yet but
there's a comment in the code about it now.
Old algorithm:
Raytrace from one transparent surface to the next step by step. To minimize
overhead in cases where we don't need transparent shadows, we first trace a
regular shadow ray. We check if the hit primitive was potentially transparent,
and only in that case start marching. this gives extra ray cast for the cases
were we do want transparency.
New algorithm:
We trace a single ray. If it hits any opaque surface, or more than a given
number of transparent surfaces is hit, then we consider the geometry to be
entirely blocked. If not, all transparent surfaces will be recorded and we
will shade them one by one to determine how much light is blocked. This all
happens in one scene intersection function.
Recording all hits works well in some cases but may be slower in others. If
we have many semi-transparent hairs, one intersection may be faster because
you'd be reinteresecting the same hairs a lot with each step otherwise. If
however there is mostly binary transparency then we may be recording many
unnecessary intersections when one of the first surfaces blocks all light.
We found that this helps quite nicely in some scenes, on koro.blend this can
give a 50% reduction in render time, on the pabellon barcelona scene and a
forest scene with transparent leaves it was 30%. Some other files rendered
maybe 1% or 2% slower, but this seems a reasonable tradeoff.
Differential Revision: https://developer.blender.org/D473
This was the original code to get things working on old GPUs, but now it is no
longer in use and various features in fact depend on this to work correctly to
the point that enabling this code is too buggy to be useful.
This can for example be useful if you want to manually terminate the path at
some point and use a color other than black.
Reviewed By: brecht
Differential Revision: https://developer.blender.org/D454
for one of the input shaders is zero.
This gives about 5% speedup for koro_final.blend. In general this is important
so you can design shaders that run faster for shadows, diffuse bounces, etc, for
example by skipping procedural textures or even using a single fixed color.
This makes it easier to have per kernel number of registers. Also, all the
tunable parameters for this are now in kernel.cu, rather than spread over cmake,
scons and device_cuda.cpp.
This fixes the ptaxs "ACCESS_VIOLATION" error and should allow our Linux and Windows build bots to compile again.
Unfortunately this comes with a performance penalty on sm_2x cards, so this is only a workaround for now. Branched Path is still globally disabled on GPU.
This now uses decoupled ray marching, and removes the probalistic scattering.
What this means is that each AA sample will be slower but contain less noise,
hopefully giving less render time to reach the same noise levels.
For those following along, there's still a bunch of volume sampling improvements
to do: all-light sampling, multiple importance sampling, transmittance threshold,
better indirect light handling, multiple scatter approximation.
This basically records all volumes steps, which can then later be used multiple
time to take scattering samples, without having to step through the volume
again. From the paper:
"Importance Sampling Techniques for Path Tracing in Participating Media"
This works only on the CPU, due to usage of malloc/free.
Similar to surfaces, this will now always scatter rather than probabilistically
scattering or not depending on the transmittance.
This also makes calculation of branched path throughput non-probalistic, which
makes thing slower too. That's to be solved by decoupled ray marching later.
This removes a few optimizations to avoid exp() calls and division, they will be
added back later, at the moment it's more important to make the code easier to
understand and refactor.
Rather use random point in each step instead of giving the steps random sizes.
Gives a bit more accurate results with large step sizes, but also convenient
convention for later changes.
These can currently be accessed by adding an Attribute node and specifying one
of those three names. A Smoke/Fire node should be added at some point to make
this more convenient.
These values might change still before the release, in particular for flame the
meaning seems unclear, it's just values in the 0..1 range. This is useful for
color ramps, but it might be good if this was also available as temperature in
kelvin so it can be plugged into the blackbody node. But I couldn't figure out
from the smoke code if or how this corresponds to a physical unit.
Here's a (quite poor) example file for a fire + smoke setup:
http://www.pasteall.org/blend/27990
These are internally stored as a 3D image textures, but accessible like e.g.
UV coordinates though the attribute node and getattribute().
This is convenient for rendering e.g. smoke objects where data like density is
really a property of the mesh, and it avoids having to specify the smoke object
in a texture node, instead the material will work with any smoke domain.
This now supports multiple steps and subframe sampling of motion.
There is one difference for object and camera transform motion blur. It still
only supports two steps there, but the transforms are now sampled at subframe
times instead of the previous and next frame and then interpolated/extrapolated.
This will give different render results in some cases but it's more accurate.
Part of the code is from the summer of code project by Gavin Howard, but it has
been significantly rewritten and extended.
