summaryrefslogtreecommitdiffstats
path: root/libdimension
Commit message (Collapse)AuthorAgeFilesLines
...
* Clean up internal raytracing API a bit.Tavian Barnes2010-01-133-13/+20
|
* Split finish_fn into diffuse_fn and specular_fn.Tavian Barnes2010-01-136-40/+76
|
* Implement reflection.Tavian Barnes2010-01-0910-57/+252
|
* Calculate transformed surface normals correctly.Tavian Barnes2010-01-091-3/+9
|
* Clamp negative values in phong shading.Tavian Barnes2010-01-091-1/+6
|
* Make recursion limit a scene parameter.Tavian Barnes2010-01-083-18/+22
|
* Modularize translucency.Tavian Barnes2010-01-081-28/+32
|
* Kill DMNSN_RENDER_OBJECTS.Tavian Barnes2010-01-082-6/+5
|
* Clean up internal raytracing API a bit.Tavian Barnes2010-01-071-60/+74
|
* Limit recursion depth in raytracing.Tavian Barnes2010-01-061-9/+20
|
* Make quality settings into flags.Tavian Barnes2010-01-062-14/+14
|
* Split up objects, finishes, pigments, lights, cameras, etc.Tavian Barnes2010-01-0610-182/+278
|
* Clean up finish combination allocation semantics.tavianator2009-12-271-15/+29
|
* Separate finishes into single-purpose finishes.Tavian Barnes2009-12-235-31/+172
|
* Fix argument order to copysign().Tavian Barnes2009-12-161-6/+6
|
* Work in sRGB to be more like POV-Ray.Tavian Barnes2009-12-073-185/+128
|
* Finish camera placement routine.Tavian Barnes2009-11-271-2/+3
|
* Fake camera semantics until I figure out the real ones.Tavian Barnes2009-11-272-0/+33
|
* Add source filename to dmnsn_error() output.Tavian Barnes2009-11-252-5/+11
|
* Fix progress calculation in raytracing.Tavian Barnes2009-11-201-1/+1
|
* Use dmnsn_new_*() rather than dmnsn_*_construct().Tavian Barnes2009-11-196-77/+72
|
* Refactor raytrace.c a bit.Tavian Barnes2009-11-191-56/+81
|
* Implement translucency with shadows.Tavian Barnes2009-11-191-10/+35
|
* Calculate .filter and .trans correctly in dmnsn_color_filter().Tavian Barnes2009-11-191-1/+5
|
* Make dmnsn_illuminate_color() perceptually uniform.Tavian Barnes2009-11-191-54/+54
|
* Translucency support.Tavian Barnes2009-11-186-49/+137
|
* Use CIE 1931 RGB for dmnsn_color_illuminate().Tavian Barnes2009-11-172-19/+94
|
* Implement phong shading.Tavian Barnes2009-11-173-12/+48
|
* Make finish callback a propper BRDF.Tavian Barnes2009-11-174-31/+52
|
* Use "diffuse" instead of "specular" for diffuse finish.Tavian Barnes2009-11-172-6/+6
|
* PGO build improvements.Tavian Barnes2009-11-131-3/+13
|
* Implement profile-guided optimization builds.Tavian Barnes2009-11-121-0/+9
|
* Make dmnsn_illuminate_color() work, with sRGB for now.Tavian Barnes2009-11-111-20/+7
|
* Raytrace one row of pixels per thread.Tavian Barnes2009-11-091-5/+2
| | | | | | The theory is that if the next ray shot is closer to the previous ray, the k-D splay tree will work better because we're more likely to hit the same object.
* Correct dmnsn_color_illuminate().Tavian Barnes2009-11-092-11/+19
|
* Use finishes.Tavian Barnes2009-11-094-65/+101
|
* Add support for finishes (BRDFs).Tavian Barnes2009-11-096-3/+129
|
* Calculate surface normals in intersection callbacks.Tavian Barnes2009-11-094-11/+23
|
* Render multiple (or zero) lights correctly.Tavian Barnes2009-11-091-1/+2
|
* Store inverse object transformation in a separate field.Tavian Barnes2009-11-092-13/+14
|
* Remove unused variable.Tavian Barnes2009-11-091-1/+0
|
* Rudimentary light/shadow handling.Tavian Barnes2009-11-094-21/+108
|
* Make dmnsn_delete_scene() delete the scene's elements too.Tavian Barnes2009-11-091-0/+17
|
* Add lights to scenes.Tavian Barnes2009-11-092-6/+12
|
* Add support for lights.Tavian Barnes2009-11-096-1/+183
|
* Make default fatal error handler nicer.Tavian Barnes2009-11-051-2/+27
| | | | If we can, only exit the current thread, and print a backtrace.
* Fix dmnsn_kD_splay_node_swallow() typo.Tavian Barnes2009-10-301-4/+4
|
* Don't flip optimized PNG writes.Tavian Barnes2009-10-301-2/+3
|
* Make sure we die on critical dmnsn_error()'s.Tavian Barnes2009-10-291-0/+2
|
* Major dmnsn_kD_splay_search() optimization.Tavian Barnes2009-10-261-19/+21
| | | | | | | | At each level of recursion, we have to go down the right branch if it exists. But if we do this before we test the current node and the left branch, we can eliminate those tests in the likely case that we find a closer object in the geometrically larger right subtree. This gives about a 2X speed improvement according to `make bench'.