This has all been very helpful guys! Thanks Sebastian, I've tried running our bone matrice updates without adding the model and just running nothing in the update callback both with the same results. Which even further reinforces the original suspicions that the hand model is being skinned on the CPU.
I only use a traversal once in the setup phase to obtain references to the bones so the traversal class is not used in the update class. Now as you guys saw in the picture in release mode I was getting upwards of 12 for the update traversal. I managed to apply the shader program exactly as implemented in the osganimationhardware example on the hand model. (was a problem with relative paths), and in DEBUG mode the update traversal now is only at around 4. Which is great, i think. It totally screws up my hand model though. Distorts it very much. In the output it mentions almost all of my bones being too low in Bone Weight and therefore not being updated. I THINK (not a lot of experience with bone weight) this is the reason behind the distortion. How do I handle the weight of my bones, in the hand model being so small? Is there a way I can force low bone weights into the palette? Or would that be a bad idea? Or maybe that not the cause of the distortion? Im a bit confused aka noob-struck on this one. Also I noticed in the osganimationhardware example that in the init() method there is a property: int attribIndex = 11; Where does 11 come from? Im not sure how this number is derived. Is it chosen at random or does it relate to the Nathan model? Trying to read through the code and understand it and learn so it doesn't distort the crap out of my hand model. Thank you! Cheers, Chris ------------------ Read this topic online here: http://forum.openscenegraph.org/viewtopic.php?p=62341#62341 _______________________________________________ osg-users mailing list osg-users@lists.openscenegraph.org http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org
