|
Linear – There is equal emphasis on all samples. Averaging stops on reaching the specified number of averages. Mainly used for analysis of stationary signals and transients
Exponential – Places the emphasis on the latest sample. The contribution of each sample towards the average is weighted using an exponentially decaying curve with a time constant equal to the number of averages, Na, times the record length, T. If a measurement is made with 0% overlap, Na ´ T is equal to the time constant. Exponential averaging is a continuous process. It is useful for analysing continuous non-stationary signals
Peak – Records the largest amplitude of each spectral line. When a new sample is included, values are compared at each frequency and the largest retained. Averaging stops on reaching the number of averages |
|