Actually, the modelview matrices I get (from CullVisitor::getModelViewMatrix() inside the operator()(Node*, NodeVisitor*) of a cull callback) do seem correct. It is just that I may not fully grasp what they actually contain at that point in the code. I thus have a couple of additional questions:
1) I have the impression they do not incorporate the viewing transform, and hence are just local-to-world matrices at that point. Could you please confirm this? 2) What would be a desirable way to retrieve the current viewing transform? Would calling Code: cullVisitor->getRenderInfo().getView()->getCamera()->getViewMatrix() be a nice way of doing it? 2) My goal is to send the previous frame's modelview and projection matrices to my vertex shader via uniforms. Would it make any difference to add such uniforms to the stateset of a Geode (via a NodeVisitor) or to add them the stateset of a Drawable (via a Drawable::CullCallback), given the fact the Drawable has been added to the said Geode? Would it affect the way uniforms are uploaded? Thanks in advance. :) Cheers. ------------------ Read this topic online here: http://forum.openscenegraph.org/viewtopic.php?p=59866#59866 _______________________________________________ osg-users mailing list [email protected] http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org
