I want to capture multiple actors. Does the calibration process need to be performed by each actor?

No, the calibration only needs to be captured by one actor as the system is calibrating the capture space.