Hi Robert, I'm not sure if the refactor will affect the way that stereo will be managed, but currently also exists auto-stereoscopic displays, which means that not only find the classical two-view stereo but many views (8 or 9 are the most common) and even in mobile devices it's also possible to find auto-stereoscopic displays.
Some time ago I developed a MultiviewNode to support this kind of displays, which basically creates a RTT camera with a grid of cameras below using a GUIEventHandler to apply the stereo displacement. Then renders the RTT to screen and uses a shader for rendering the specific interlave of the display. I knew that this wasn't the best approach but I needed to change the projection and view matrices to get the correct stereo result and as far as I knew this only was possible using the SceneView callbacks but not the in slave views, or at least it wasn't possible in an easy way :). I'm just commenting this because I think with the new approach will be worth to consider this kind of displays. My two cents and thank you for the efforts! :) Rafa. 2013/3/29 Robert Osfield <[email protected]> > Hi Christian, > > > On 29 March 2013 13:30, Christian Buchner <[email protected]>wrote: > >> >> Would supporting barrel predistortion for the Oculus Rift VR headset >> touch this area of work as well? Mathematically speaking it's slightly more >> advanced than keystoning, but both could be implemented with a distortion >> mesh. >> > > In principle it would work in exactly the same way, the only difference > would be the mesh/texture coordinates you'd use in the second pass, the RTT > pass would be the same. > > Given this the principle of general distortion correction would come down > to users provide the appropriate distortion scene graph and setting > projection/modelview matrix for the distortion correction Camera that > renders to the final screen. For stereo on separate display outputs we'd > have two such distortion correction settings, but for stereo on a single > display (quad buffer/anaglyphic/interleaved) then we'd have just one. > > Robert. > > _______________________________________________ > osg-users mailing list > [email protected] > http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org > > -- Rafael Gaitán Linares CTO at Mirage Technologies S.L - http://www.mirage-tech.com gvSIG3D Developer - http://gvsig3d.blogspot.com
_______________________________________________ osg-users mailing list [email protected] http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org
