Hi,
I currently do a car sim program, and I encount a problem about heigh memory &
CPU used when I add RTT camera:
I want add 3 back mirrors to the car, so I create 3 relative camera to render
the scene to a texture, and use root->addChild(mirror_camera) to bind camera to
root node, these mirror camera render the same scene but with different
shader(which is more simple than main camera rendering), the struct of the root
is like this:
root
/ \
mirror_camera0 group(complex shader)
/ \
group(bind simple shader) scene_node
/
scene_node
the scene is load from ive file which is about 10MB, when I render with no
mirror camera, the memory left is about 900MB( I read from windows taskManager
and the total memory of my computer is 2GB), and when I use 3 mirror cameras,
the memory left is only about 80MB, and also high CPU used. and My scene is
very simple which only contained 2 textures and 80 Geometry, with about 400,000
tri_strip, In my opinion the scene_node can be shadered between diffferent sub
cameras and main scene, is there anything wrong with my approach?
Thank you!
Cheers,
Kai
------------------
Read this topic online here:
http://forum.openscenegraph.org/viewtopic.php?p=28769#28769
_______________________________________________
osg-users mailing list
[email protected]
http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org
