A phon loudness model quantifying middle ear and cochlear sound compression

Abstract

Phon loudness data by Lydolf and Mo/ller [1997, see Suzuki and Takeshima, J. Acoust. Soc. Am. 116, 918 (2004)] at third octave frequencies 20–1000 Hz suggest evidence for Middle-Ear (ME) Sound Compression (MESC) at loudness above 60 phons. Parameters of a phon model based on sound compression by an idealized cochlear amplifier (Goldstein, 2009, 2011) were fit to the phon data at low sound levels. MESC was estimated as the excess in sound pressure levels of phon data over model predictions for 80–100 phons. MESC was found when cochlear amplifier inputs exceeded the known Acoustic Reflex Threshold (ART) of ~83 dB SPL, causing ME Muscle (MEM) contraction. Physiologically consistent first-order highpass ME transfer functions reveal a mix of MEM attenuation and attenuation at 20–100 Hz attributed to ME static pressure from pressure-field stimuli >100 dB SPL. These attenuations are isolated with the ME modeled as the low side of an RLC analog bandpass. The original data above 60 phons are modeled accurately as the sum of the two attenuations, the cochlear amplifier input, and normal ME attenuation re 1 kHz. The considerable basic information estimated from the phon data can be compared with other sources to guide further integrative research. [NIH funded 1972-04.]