Blender Git Commits

Blender Git "render25" branch commits.

Page: 28 / 29

February 23, 2010, 19:48 (GMT)
Render Branch: Irradiance Cache

Implementation borrows some code from Pixie though has changed quite
a bit. This is work in progress and has artifacts and hardcoded
parameters still.

I tried irradiance gradients and found them problematic, they seem
to reduce part of the artifacts but for others have no effect, and
can even make things worse. There's a clamping trick that helps a
bit here, but I couldn't get that reliable either.

Currently it is using least squares reconstruction to smooth out the
results, it's too smooth now, still trying to find out the best way
to weight samples without introducing problems, but I'm finding this
much more promising than gradients.

February 23, 2010, 19:34 (GMT)
Render Branch: quick hack to support 4x4 matrix SVD decomposition
in C, by using the TNT matrix library from the IK solver.

February 23, 2010, 19:27 (GMT)
Render Branch: math functions

* add len_squared_v3v3
* add mul_v4_m4v4
* add box_minmax_bounds_m4
* rename disc_to_sh to vec_fac_to_sh

February 23, 2010, 19:24 (GMT)
Render Branch: redrawing while render would miss a few scanlines each time
with antialiasing enabled.

February 23, 2010, 19:23 (GMT)
Render Branch: disable ray trace epsilon for testing, not using it
gives artifacts with overlapping faces, but it also seems to lead
to self intersections.

February 11, 2010, 19:29 (GMT)
Render Branch: WIP commit

* Pixel cache now works for raytraced AO/env/indirect.
* Falloff distance value is now only used for custom curve, linear and
quadratic now have a smooth factor (range 0-1).
* Lamp multisampling code moved from shadow to shading, so that both
shading and shadow can be multisampled. This should give correct area
lights at the cost of noise.
* Halo now uses same lamp code as surface instead of duplicating it.
* Fix in volume code, was too bright now with lamp changes.
* Only shadow materials are now computed different, sharing more code
but not matching old behavior yet.

February 11, 2010, 19:20 (GMT)
Render Branch: copy 2nd order spherical harmonics code from render to math lib.

February 10, 2010, 14:54 (GMT)
February 5, 2010, 12:45 (GMT)
Render Branch: move pixel & surface cache code to own file for reuse,
not really generic yet but getting there.

February 4, 2010, 21:00 (GMT)
Render Branch: some lowres geometry raytracing test code, just a test
that is far from usable but might as well commit it.

February 4, 2010, 20:48 (GMT)
Render Branch: stupid path tracer I'm using for consistency testing,
only does environment lighting + diffuse at the moment, don't expect
this to make it into trunk.

February 4, 2010, 20:41 (GMT)
Render Branch: lamp falloff

* Change lamp falloff formula which was dist/(r^2 + dist) with dist
a user specified value, whereas the correct value is 1/(r^2). Now
it is 1/(r^2 + dist) with dist = 0 as the default.
* Deprecated the Lin/Quad weighted mode.
* Still more work needed to make sun/area/hemi work better with this,
results are too bright now.
* Also this gives quite harsh results when the lamp is near geometry,
perhaps because we are not using tonemapping by default.

February 4, 2010, 20:33 (GMT)
Render Branch: fix memory leak when cancelling preview render.

February 4, 2010, 17:38 (GMT)
Render Branch: different method to compute only shadow lamps, rather
than subtracting colors it now subtracts in grayscale, helps avoid
ugly discolorations when used with colored lamps.

February 4, 2010, 17:17 (GMT)
Render Branch: trying to get units figured out for materials and lamps,
see comments in material.h and lamp.h. Change in UI is that lamp energy
is now called power and is specified in Watts. It's still off by a large
factor due to the lamp falloff code, that will be fixed in a separate
commit.

February 4, 2010, 16:55 (GMT)
Render Branch: temporary workaround for extern texture access, previously this
was using the global R, with whatever contents it had from last render. For now
just introduce a static local R for this case.

February 4, 2010, 16:53 (GMT)
Render Branch: make the raycounter code work again, broke it in refactoring.

February 4, 2010, 16:35 (GMT)
Render Branch: fix for sampler.c, hemisphere sampling code was wrong,
also added back cosine weighted hemisphere.

January 28, 2010, 22:23 (GMT)
Render Branch: fixes for crashes in bake and occlusion caching.

January 28, 2010, 16:02 (GMT)
By: Miika HämäläinenLast update: Nov-07-2014 14:18MiikaHweb | 2003-2021