We’re exporting a scene from maya that has a number of static camera cuts. When the cameras are exported to collada each cameras visualScene has a node and a transform matrix for the node. Because the scene is unanimated, this transform is used to set the position and orientation of the static camera for that cut.
The problem is that the node transform is the same for each camera, it always is set to the transform of the camera in the very last frame of the scene rather than the transform of the camera involved in that cut. Is this a known bug in the collada exporter, or is there is something that hasn’t be exported correctly?
Could you let me know which maya plug in that you are using? Feeling one or Net Allied one? And could you share your Maya file with me then I can take a look at that.
More specific: there a two different COLLADA Maya implementations on sourceforge: one is called ColladaMaya_FREE_3.05B and is the old one. The actively maintained one is called COLLADAMayaNextGen. Please try both of them and post your results on the Bugtracker on sourceforge at COLLADA for Maya, 3DSMax download | SourceForge.net, too.
I don’t think that I can share the Maya file because of NDA’s and such things. I’ll check out the exporter to see if that exports the transforms correctly and get back to you