Hi Robert,
> The above call *is* required for the pre RenderStage, as this can
> have their own pre RenderStages.
Aaahh, I see. Now this begs the question, why is SceneView calling it
explicitly? Is it because certain stereo modes, like anaglyphic,
change the the global state, and the pre render stages need to be
called before this?
When you render anaglyphic and are using RTT via the frame buffer you have to make sure
that all the RTT stages are done before either eye's main stage can be done.
> It depends on what you are trying to do. If you want a single result for
> both camera's then you just need one CameraNode, if you want seperate
> results then you'll need to create two CameraNode's one for each seperate
> eye and use NodeMask/TraversalMask to select the appropriate one.
Is this by design or a side effect of the current implementation?
Basically, would you be against a patch that would allow a single
CameraNode to work in stereo?
Its not a case of making CameraNode work in stereo, its how you manage everything at a higher level. Stereo certainly complicates things and there is any one right answer. However, as an application developer you know more about the domain you are working in so better make decisions about which answer works best for you, which is why the NodeMask/TraversalMask approach is useful - it puts the solving the problem in your hands.
Robert.
_______________________________________________ osg-users mailing list [email protected] http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
