In recent years the software coincidence counting system, designed for absolute activity measurement, has been developed in the Czech Metrology Institute. In this system a true coincidence count rate is calculated from the records of time and amplitude data of individual pulses and may be determined by two different methods. The first one uses a coincidence resolving time, in a manner similar to a classical coincidence measurement. The second method applies the pulse mixing method formulae, so that it does not use the resolving time and the correction for accidental coincidences. Both methods have been tested on the same data from a 4 πβ (PC)- γ coincidence measurement of 60Co sources. The difference between the results obtained from both calculation methods applied on the data from the same measurement did not exceed 0.015%. The details of both methods and the results of their comparison are presented.