On Wed, Aug 6, 2008 at 3:06 PM, hesicong2006 <[EMAIL PROTECTED]> wrote: > Thanks Robert, I'll have a try! > > Can I ask another question not relative to OSG? > > I now encounter a performance issue of volume rendering. With the increasing > size of 3D texture (256*256*256*2, Lumance and Alpha), the speed of ray > casting algorithm slows down very significant. I guess it will due to the > long texture fetching and filtering of large 3D texture. So I have to split > the large 3D texture use small 3D texture( 32*32*32 or 64*64*64), render > small boxes and compose them. By doing this, the texture cache on the card > will speed up the texture fetching. I don't know if it is true. Thanks first > if someone can give me a hand!
There is no straight answer to this question as the answer is hardware/driver dependant. In general I'd say small bricks should produce better cache coherency, until I've tested it out personally with volume rendering I can only speculate. Robert. _______________________________________________ osg-users mailing list [email protected] http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org
