On 7/23/2012 7:14 AM, Ron Mayer wrote:
When the application comes in with a delta [x, y], we only move the frustum of the orthographic projection, without moving the camera. I would imagine the right thing to do would be to move the camera in [x, y] without changing the frustum. But for some reason just moving the frustum works fine. I am not sure if this is valid or not. [I do know that if we were in perspective we would have to move the camera, since the frustum is defined by the angle and ratio of the view.]
I suspect that the concatenation of the modelview and projection matrices are the same in either case, so the end result of the transformation is the same.
But it's likely that you really do want to effect the movement in the modelview matrix rather than the projection matrix, and the reason for this is that OpenGL performs certain operations in eye coordinates, and the movement would need to be in the modelview matrix in order for these operations to be performed correctly. The operations including lighting and clip planes.
All this is covered extensively in books like OpenGL Programming Guide and (shameless plug) OpenGL Distilled.
-Paul _______________________________________________ osg-users mailing list [email protected] http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org
