Multicomponent acoustic distortion product otoacoustic emission phase in humans. II. Implications for distortion product otoacoustic emissions generation.

Phase characteristics and latency of 2f1-f2, 3f1-2f2, and 2f2-f1 acoustic distortion product otoacoustic emissions (DPOAEs) recorded in normally hearing adults have been described in a companion paper [Moulin and Kemp, "Multicomponent acoustic distortion product otoacoustic emission phase in humans. I. General characteristics," J. Acoust. Soc. Am. 100, 1617-1639 (1996)]. Phase-versus-frequency functions (PFF) have been recorded using two methods, depending whether f1 or f2 was swept in frequency, defining, respectively, an "f1 sweep" and an "f2 sweep" method. The present study compares 2f1-f2, 3f1-2f2, 4f1-3f2, and 2f2-f1 DPOAEs phase characteristics, and latencies obtained by these two methods. For lower sideband DPOAEs, the f2 sweep gave consistently longer latencies than the f1 sweep method, whereas no difference was obtained for the 2f2-f1 DPOAE. This suggests that the lower sideband DPOAEs and the 2f2-f1 DPOAE do not come from the same aspect of the traveling wave. The longer delay obtained with an f2 sweep is suggested to be due to an additional delay added by the proximity of the peak of the f2 traveling wave. The place-fixed and the wave-fixed models of DPOAE generation are discussed, and an "intermediate" model of DPOAE generation is introduced to explain such additional delay and the existence of phase irregularities which were observed in the PFF. The ratio of the latencies obtained by the two sweep methods is frequency dependent and a relation to cochlear frequency selectivity is suggested. Finally, analysis of the differences in latency of the different lower sideband DPOAE components allowed an estimation of the return latency of DPOAEs.