Okay, I've been burrowing down into some of the accuracy problems in OSG with picking POINT_ROT_EYE billboards via View::computeIntersection. I fixed one of the problems, but now my brain has turned to mush, so I wanted to see if anyone had some insights on how we could fix the other issue I'm having...
Right now, if you do computeIntersections() with a screen location, it does a line segment intersector from the near clip to the far clip at that location. That's correct, but the reference eyepoint is not -- it's set to the start point on the near clip plane. That's not right for this situation. You want the billboards to still face the original eye point. Unfortunately you cannot provide the original eye point if you're using the WINDOW or PROJECTION coordinate frame -- it's at infinity. Maybe we need to add some special logic to handle WINDOW/PROJECTION coordinate frames in the Billboard routines of IntersectionVisitor? Any thoughts? ------------------ Read this topic online here: http://forum.openscenegraph.org/viewtopic.php?p=29647#29647 _______________________________________________ osg-users mailing list [email protected] http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org
