Do the method(s) used to blend shapes change between static and dynamic modeling? How is this change accounted for and does it differ in general depending on desired application and outcome?