Spontaneous otoacoustic emissions: incidence and short-time variability in normal ears

More than 50% of normal ears generate spontaneous otoacoustic emissions, which are sinusoidal tones originating in the cochlea. These faint sounds with levels typically at 10 dB to -30 dB are usually not recognized by the subjects who have them, but can be recorded by sensitive microphones in the ear canal. The clinical significance of having or not having spontaneous emissions are at present unknown. The emissions may be useful for non-invasive monitoring of cochlear physiology. A total of 307 spontaneous otoacoustic emissions from 59 ears have been examined. The short-time variability was small when as many as 300 averages were collected. High-frequency emissions showed better stability than low-frequency ones. The ratio of frequency distance between neighboring emissions to critical bandwidth (Bark) was preferentially 0.3-0.5 for lower frequencies and 0.2-0.4 for the higher ones. Amplitude dependence for one emission on its neighbors was difficult to demonstrate. Combination tones (different tones f2-f1 and the cubic distortion product 2 x f1-f2) are probably rare among spontaneous emissions.