Ear canal acoustics was examined using a onedimensional lossy transmission line with a distributed load impedance to model the ear. The acoustic input impedance of the ear was derived from sound pressure measurements in the ear canal of healthy human ears. A nonlinear least squares fit of the model to data generated estimates for ear canal radius, ear canal length, and quantified the resistance that would produce transmission losses. Derivation of ear canal radius has application to quantifying the impedance mismatch at the eardrum between the ear canal and the middle ear. The length of the ear canal was found, in general, to be longer than the length derived from the one-quarter wavelength standing wave frequency, consistent with the middle ear being mass-controlled at the standing wave frequency. Viscothermal losses in the ear canal, in some cases, may exceed that attributable to a smooth rigid wall. Resistance in the middle ear was found to contribute significantly to the total resistance. In effect, this analysis "reverse engineers" physical parameters of the ear from sound pressure measurements in the ear canal.

Abstract. The ear canal, below about 6 kHz, is well described by a uniform cylinder (sound propagates predominantly as plane waves) with the middle ear being a non-rigid termination. A nonrigid termination can be viewed as altering, as a function of frequency, the acoustic length and radius of the cylinder. It is generally assumed that sound transmission in the ear canal over this frequency range is lossless. This paper presents a method for calculating the influence of visco-thermal losses and the middle ear on ear canal acoustics. The acoustic input impedance was derived from sound pressure measurements in the ear canal and then a nonlinear least-square-fit to the data with a one-dimensional model incorporating visco-thermal losses generated length, radius, and middle ear impedance parameters. It was found that a rigid wall assumption for visco-thermal calculations was insufficient to account for damping in the ear canal. The properties of the ear canal wall (not being a rigid, low-friction surface), incorporated into visco-thermal losses as a scaling factor, provided a better fit to the data. Viscous and thermal losses were both found to affect sound propagation in the ear canal, viscous losses being more significant, altering the acoustic input impedance of the ear primarily in the region of the standing wave frequency. The model data suggests that the middle ear influences ear canal acoustics up to about 3 kHz.