Voice Stress Analysis

The Promise of Voice Stress Analyzers

The promise of a better lie detector, one that can be used covertly, is what has been offered by developers of various so-called voice stress analyzers (VSAs). These devices all claim to be able to detect lies by electronic assessment of certain frequency changes in the vocal spectrum. They have been known by various brand names, such as the initial device marketed from about 1970 to the mid-1980s, the Psychological Stress Evaluator (PSE) and its imitators—Hagoth, Verimetrics, LEA Mark V, and the Computerized Voice Stress Analyzer (CVSA).

The basic “theory” on which all of these devices are based is that, in the spoken word (i.e., the vocal spectrum), there are subaudible, involuntary frequency modulations (FM) that are superimposed on audible voice frequencies. The strength and pattern of these frequency modulations vary inversely with the degree of stress in a speaker at the moment of an utterance, such as “yes” or “no” or more lengthy statements. It is claimed that the FM results from minute oscillations of the muscles of the voice mechanism; these are known as physiological tremors and are said to be under the control of the central nervous system (i.e., the voluntary system) during nonstressful periods, but as stress is imposed, the autonomic nervous system (generally, involuntary) gains dominance, resulting in a suppression of the microtremors. This suppression, said to indicate stress, is detected and displayed by voice stress analyzers in characteristic blocked or rectangular wave forms, which, moreover, are indicators not only of stress but also, in their shape, of the degree of stress being experienced.

...