Hi Tom, The CPU cost of depth sorting the transparent objects is pretty low, and will only be significant if you have a very large number of geometries that need to depth sorted.
Do you think you are CPU limited by the sorting of transparent bin? Robert. On Wed, Oct 27, 2010 at 11:19 AM, Thomas Hogarth <[email protected]> wrote: > Hi all > I'm trying to push the IPhone fps to the limit, at moment I can render > 10,000 + triangles with diffuse/normal/reflection mapping at 30 fps but I > think we can squeeze a lot more out of it. So i'm writing a node visitor > which could eventually be added to the optimizer to remove all un used > states etc when running on gles 2. This is the majority of the fixed > function pipeline like texture matrices etc. > On top of this IPhones and perhaps all devices using the PowerVR chips use > a deferred renderer, which means depth sorting is pointless as the renderer > will wait till all commands are received and ensure they are rendered in the > correct order anyhow. So my question is is the best way to disable depth > sort to put everything into the transparent bin? > Also if anyone has anyother suggestions to squeeze a few more fps out of > these devices I'd love to hear them. > Cheers > Tom > _______________________________________________ > osg-users mailing list > [email protected] > http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org > > _______________________________________________ osg-users mailing list [email protected] http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org
