Abstract
Hans Kunov reports (Kunov, 1977) that non-linearities in the eardrum area severe obstacle to the recording of the ipsilateral response of the acoustic middle ear reflex. His paper concerns commercial equipment having a probe tone frequency of 220 Hz. The obstacles mentioned by Kunov are related to the vibration amplitude of the eardrum and consequently are brought about by the very choice of probe tone intensity and frequency. It has been shown earlier that a probe tone frequency of about 800 Hz is optimal because it will allow the highest signal-to-noise ratio to be obtained in the recordings of acoustic impedance changes of the middle ear (Møller, 1961a; b). In the following comments it will be shown that such artifacts as reported by Kunov do not exist when the frequency of the probe tone is in the region of 600–1000 Hz, where a lower sound level of the probe tone gives the same or an even better signal-to-noise ratio compared with a probe tone frequency of 220 Hz (Møller, 1961a; 1962). The reason that a 800 Hz probe tone gives an optimal signal-to-noise ratio is that this frequency is close to the middle ear's principle resonance frequency. The acoustic impedance in the plane of the eardrum has its lowest value at 800 Hz. This impedance is in parallel with the impedance of the air volume in the ear canal and the impedance of the measuring device. A change in these impedances will then have its least relative influence on the total acoustic impedance in the frequency range near the principle resonance frequency of the middle ear. A certain change in the ear's impedance at that frequency will thus affect the total measured impedance to a greater extent than changes at lower frequencies (Møller, 1960; 1961a). These considerations were taken into account in developing the apparatus shown in Fig. 1 (Møller, 1961a; 1962). With this set-up, ipsilateral change in impedance as well as contralateral changes can be recorded without artifacts for stimuli within a large range of intensities. Typical recordings using the set-up shown in Fig. 1 and two probes, one in each ear simultaneously, are illustrated in Fig. 2. The results concern a human subject with normal hearing. The stimulus frequency was 1450 Hz (Møller, 1962). Noise stimuli have also been used to elicit the stapedius reflex with the same set-up. No artifacts arose, as can be seen from Fig. 3, where the responses to 1450 Hz tones are compared with those to bandpass filtered noise, centred at 1450 Hz and with band-widths of 300, 200 and 100 Hz. Fig. 4 shows the stimulus response curves for bilateral (dashes and dots), ipsilateral (solid line) and contralateral stimulation for different frequencies. Fig. 5 shows similar stimulus response curves for tones and noise.