Modeling the fine structure of the 2f1-f2 acoustic distortion product. II. Model evaluation.

In a previous article, a vector sum model was developed that successfully reproduces the ADP rippling pattern when a nonuniformity is introduced in the active damping factor, a parameter that is inversely related to the energy gain contribution of the outer hair cells [OHCs, Sun et al., J. Acoust. Soc. Am. 96, 2166-2174 (1994)]. Here, the mean of the damping factor is increased nonlinearly with input level, mimicking the saturation of the active feedback of the OHCs. The passive damping factor in the transmission line model was also nonlinearly increased with input level to reproduce the frequency shift of the peak of the traveling wave observed in experimental data. The resulting model simulates an ADP that is compatible with data from human subjects wherein the ADP fine structure does not saturate with level. Moreover, the model suggests that the shifting of the ADP pattern with level is a direct result of the peak shift of the traveling wave, thus implicating the nonlinear damping factors as the underlying basis of this phenomenon. The input/output (I/O) functions of the simulated ADP emissions at specific frequencies were also examined. The resulting functions show a variety of shapes, depending on the pattern of ADP fine structure around the I/O frequency, and the way the fine structure shifts as primary levels increase. These I/O functions are also similar to those observed in human subjects, even with regard to overall slopes which approximate one. Thus the model illustrates how cubic distortion generators coupled with damping on linearities can yield I/O function slopes on the order of one, rather than the expected three.(ABSTRACT TRUNCATED AT 250 WORDS)