Frequency discrimination of random-amplitude tones.

The ability of human observers to discriminate frequencies of tones between 1000–15 000 cps was measured in a temporal two‐alternative forced‐choice discrimination experiment. On each trial, one of the test tones, selected at random, was attenuated by a random amount ranging from 8 to 20 dB in 2‐dB steps. Comparison of the results of this experiment with those of previous experiments in which fixed‐amplitude tones were used, indicates little difference between the observers' abilities to discriminate frequencies of fixed‐ and random‐amplitude tones at 1000 cps. Discrimination with random‐amplitude tones becomes relatively poorer, however, as frequency is increased. For example, the classical data show a frequency just‐noticeable difference (jnd) of 40 cps at 10 000 cps; in the present experiment, with random‐amplitude signals, the observers were unable to achieve 75% correct responses until ΔF was 300 cps.