Hi Paul, Aha I see you've looked at the sandbox code.
Yes your understanding about the vertices is correct. As for the normals I'm simply going to compute the normals for the high resolution set and then use them for the lower resolution subsets too. I've tried this with the normals before and even though the normals are technically incorrect for the new lower resolution primitive sets you would be very hard pushed to notice the difference. Basically I plan to implement this as an alternative OceanTechnique and keep the existing brute force approach for legacy apps. I'm also planning to abstract the FFTSim class out so that it can be modified more easily paving the way for alternative methods of generating the ocean tiles. Let me know if you have any other queries. Kim. On 16 February 2010 21:22, Paul Palumbo <[email protected]> wrote: > > > > A little while ago I wrote some test cases which > > > removed this huge vertex array and replaced it with a single high > > > resolution ocean tile which is then instanced for each tile in the > > > ocean surface. The different resolution tiles are then supported > > > simply by modifying the primitives and positioning the tiles using a > > > vertex shader. Since all of the high resolution tiles are precomputed, > > > updating the animation of the ocean surface is simply a matter of > > > changing the pointer to the next frames vertex array for each of the > > > tile primitives. > > > > Do I understand this correctly? Only the "high resolution" vertices will be > used and the geometry will simply reference the appropriate subset of > vertices based on the level of detail? Therefore, only 1 OceanTile object > would be needed (for the highest level) and the other levels-of-detail > wouldn't be needed. Then the MipmapGeometry::updatePrimitives call would > just update the geometry to references the "appropriate" high-level" > vertices based on the current resolution. > > I assume the same would be true for the Normals where only the normals for > the high resolution would be used. Your TilePrimitives example kind of > brushes over the Normals and the existing OceanTile::computeNormals seems to > be resolution dependent. > > Do I have this correct? Just trying to understand what you are planning. > > Thanks > Paul P. > > ------------------ > Read this topic online here: > http://forum.openscenegraph.org/viewtopic.php?p=24248#24248 > > > > > > _______________________________________________ > 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
