On the Power Gained by Clipping Audio and Single Side‐Band Speech Signals

Available statistical distributions of instantaneous speech amplitudes have been examined, and an arbitrary but reasonable peak factor has been chosen for a clipping reference level. The power increase in the audio wave (and hence the AM or DSB side bands) due to peak clipping in the audio band has been calculated by using the standard deviation of the clipped and unclipped distributions as a measure of the root‐mean‐square voltage. A curve will be presented showing the power gain as a function of clipping, and the relation of this to the intelligibility will be pointed out. Statistical distributions of the envelope of a single side‐band speech signal have been run and compared with those taken by others, and these have been treated in a manner similar to the audio case to determine the power gained by clipping the single sideband signal. A curve of the power increase under these conditions will also be presented. Unfortunately, there are no data on the intelligibility of clipped single side‐band speech to which the computed curve may be related.