Hehe thanks guys, the plush toy was just a test piece, we'll eventually be looking to do food shots for a client, ie see the banana below that was shot by walking around the banana (mind you that was only with 30 photos, so it was a rough test). The client wants about 400 food shots done, which we're now thinking a QTVR approach might be better for time's sake.
But for the future, masking and filling the frame should fix the software's camera positioning issues?