Hawkins and Stevens revisited at low frequencies
Adrianus J. M. Houtsma · The Journal of the Acoustical Society of America · 1998
The classical data on masking of pure tones by white noise [J. E. Hawkins, Jr. and S. S. Stevens, J. Acoust. Soc. Am. 22, 6–13 (1950)] yield a nonmonotonic critical ratio function that increases at low and at high frequencies. Modern bandwidth estimates of auditory filters, however, yield equivalent rectangular bandwidth (ERB) functions that are monotonic with frequency. This apparent discrepancy has sometimes been interpreted as evidence for frequency-dependent detection efficiency in the auditory system. Another possible reason for the discrepancy, however, could be the fact that the classical data were measured with headphones that needed substantial correction at low frequencies to compensate for acoustic leakage. To test this, the Hawkins and Stevens experiment was repeated with Etymotic ER-2 insert phones over a frequency range from 30 to 8000 Hz. An adaptive forced choice, as well as a Bekesy tracking procedure, was used. The resulting critical ratio function shows monotonic dependence on frequency, down to 90 Hz, similar to the ERB function. This appears to make critical ratio data, in principle, consistent with ERB data measured with notched-noise techniques, and largely eliminates the need for a frequency-dependent detection efficiency in the auditory system.