This technology is used in the Quattro3 and the Octopus. This means that the unit will assess the signals picked up by all the microphones, and decide instantaneously which one(s) will take part in the mixing and with what level of emphasis.