engine: implements the capacity for tsstatic objects to allow themselves to use the reflector system in order to display runtime generated cubemaps.
script: defines a calibrated reflectordesc that reflects all object types within the space of approximately half stock veiwdistance
usage: cubeReflectorDesc = DefaultCubeDesc placed in a given object-insatnces entry that uses a material with a dynamiccubemap = true; flag.
immediate purpose: consistency of application of materials.
long term impact: tags steps required in order to associate a given runtime generated cubemap with an object-instance. in the future, this is likely to give way to an area-derived cubemap based on the accumulation volume tech developed by Andrew Mac, or a screenspace reflection.
- Fix for issue https://github.com/GarageGames/Torque3D/issues/525 This
fix takes into account the skewed view into the world when you have a
projection offset and the ability to see further into the scene at the
edges opposite to the offset.
- SceneCullingState now has two frustum rather than one: a culling
frustum and camera frustum.
- The camera frustum should be referenced when you need the projection
matrix or don't want a skewed frustum.
- The culling frustum should be referenced during any scene culling or
when determining what dynamic geometry to render. It currently skews
itself to take into account any projection offset (automatically
calculated in SceneCullingState constructor).
- When there is no projection offset, the camera frustum and culling
frustum are the same. This usually means any time when not using the
Oculus Rift.
Removing the "using namespace Torque" from header file to prevent issues
with certain engine addons conflicting with ::UUID defined by windows
and Torque::UUID defined by T3D.
