Joerg H. Mueller
|
412d40f8c0
Moving draw item related classes to MultiDevice.
|
2 years ago |
Joerg H. Mueller
|
8048e50019
Renaming classes with MultiDevice equivalent to SingleDevice.
|
1 year ago |
Alex Peterson
|
d88545929a
Fix merge error in Culling.cpp
|
2 years ago |
Guthrie Adams
|
81fe2d94e4
Optimize masked occlusion culling to only clear the buffer if something visible was drawn to it. This keeps it from clearing in scenes that have no occlusion planes, which many scenes don't.
|
2 years ago |
gadams3
|
fe42e03123
removed Unicode quotation marks
|
2 years ago |
gadams3
|
f713979139
Updated comments and added early return
|
2 years ago |
Guthrie Adams
|
06a8d680e5
switching test culling return to default to true in case all systems are disabled
|
2 years ago |
Guthrie Adams
|
daebc84d8a
fixing gradle compiler warning as error
|
2 years ago |
Guthrie Adams
|
6071548d13
updating occlusion culling related changes after PR feedback
|
2 years ago |
gadams3
|
968afeb725
Creating occlusion request bus as interface to support alternate occlusion calling implementations. The interface supports creating occlusion views of scene data that can be used to make occlusion queries for entities and bounding boxes. An implementation should supports requests in parallel to each occlusion view.
|
2 years ago |
moudgils
|
7841fb98f6
Shadow cascade pass optimization by rejecting more meshes based on frustum intersection between light and camera. (#16032)
|
2 years ago |
Ken Pruiksma
|
eaf725f8f4
Directional shadow exclusion frustums (#15753)
|
2 years ago |
Tommy Walton
|
d2d54140aa
Add support for VisibleObjectLists in the RPI Culling. (#15517)
|
2 years ago |
JLeclercq
|
5edacbe24b
Atom/RPI: Don't compute screen coverage if lod is fixed (#13606)
|
3 years ago |
Ken Pruiksma
|
0ed5937e2d
Cached projected shadows (#13288)
|
3 years ago |
Ken Pruiksma
|
833532c74f
Light type culling with shader options (#12456)
|
3 years ago |
antonmic
|
d001640f0c
compile fix for culling debug ifdef
|
3 years ago |
antonmic
|
d31976e147
final culling performance improvements
|
3 years ago |
antonmic
|
61f2d5101d
further culling perf improvements - culling now down to ~1.5ms using older version of RosCon Demo as a stress test
|
3 years ago |
antonmic
|
941e3e47cf
Improved culling performance by 2.5ms (roughly 50%)
|
3 years ago |
antonmic
|
ac4bfc046d
Improved CPU performance by almost 13ms (37.48ms -> 24.74ms) on the RosCon level
|
3 years ago |
Ken Pruiksma
|
c0e891a190
Support frustum shape in AuxGeom (#11728)
|
3 years ago |
Jeremy Ong
|
04147eaccc
Add cstr labels to all task graphs and events
|
3 years ago |
santorac
|
bc30158e17
Made ImGuiCpuProfiler include averages for the high-level statistics, and a button to reset the collected averages.
|
3 years ago |
rgba16f
|
c9b814d4f9
Update runtime switch for using TaskGraph to default true (#8492)
|
3 years ago |
Jackerty
|
e03d9eefd9
Moves Culling.cpp::ProcessWorklist debug variables under preprocessing (#7043)
|
4 years ago |
Nemerle
|
d8e06d9d95
Apply reviewer suggestions
|
4 years ago |
nemerle
|
5ff65be314
This reduces non-unity build time by ~2% and build size by ~0.5%.
|
4 years ago |
sphrose
|
1a94caee47
Fix debug compilation.
|
4 years ago |
Esteban Papp
|
5cca52e73c
Merge branch 'development' into redcode/driller_removal
|
4 years ago |