EFFECT OF TAPE-RECORDING ON PERTURBATION MEASURES

Tape recorders have bean shown to affect measures of voice perturbatio n. Few studies, however, have been conducted to quantitatively justify the use or exclusion of certain types of recorders in voice perturbat ion studies. This study used sinusoidal and triangular waves and synth esized vowels to compare perturbation measures extracted from directly digitized signals with those recorded and played back through various tape recorders, including 3 models of digital audio tape recorders, 2 models of analog audio cassette tape recorders, and 2 models of video tape recorders. Signal contamination For frequency perturbation value s was found to be consistently minimal with digital recorders (percent jitter = 0.01%-0.02%), mildly increased with video recorders (0.05%-0 .10%), moderately increased with a high-quality analog audio cassette tape recorder (0.15%), and most prominent with a low-quality analog au dio cassette tape recorder (0.24%). Recorder effect on amplitude pertu rbation measures was lowest in digital recorders (percent shimmer = 0. 09%-0.20%), mildly to moderately increased in video recorders and a hi gh-quality analog audio cassette tape recorder (0.25%-0.45%), and most prominent in a low-quality analog audio cassette tape recorder (0.98% ). The effect of cassette tape material, length of spooled tape, and d uration of analysis were also tested and are discussed.