Antiphasic Conditions Research Articles

Binaural Lateralization of Cophasic and Antiphasic Clicks

Four subjects listened binaurally to equal-amplitude, 100-μsec rectangular pulses. The pulses were delivered at a sensation level of 40 db and at a rate of 10 sec−1. For wide-band reproduction and good transient response, the stimuli were transduced by condenser microphones fitted with ear insert plugs. The relative times of arrival of the pulses at the two ears (both lead and lag) were under continuous control of the subjects. The latter adjusted the interaural time difference to produce fused sound images. The polarities of the pulses were connected for cophasic (rarefaction-rarefaction) and antiphasic (condensation-rarefaction) conditions. Besides the unmasked conditions, lateralizations were made when the two ears were masked by additive, uncorrelated noise, high-pass filtered at 600, 1200 and 2400 cps, respectively. The results indicate that: (a) the interaural time difference for principal fusion of the cophasic condition, both masked and unmasked, is nominally zero; (b) the interaural time difference for principal fusions of the unmasked, antiphasic condition is on the order of 200 to 300 μsec; (c) masking by uncorrelated noise, high-pass filtered at frequencies less than about 1–2 kc/sec, produces antiphasic lateralizations at time differences of approximately ±1/2fc, where fc is the cutoff frequency of the noise; (d) masking by noise high-passed at frequencies greater than 1–2 kc yields lateralizations essentially the same as the unmasked case. Secondary fusions, generally lying off the medial plane, are frequently heard. Earlier calculations on the mathematical models of basilar membrane displacement [J. Acoust. Soc. Am. 32, 1494(A) (1960)] suggest the present results and, in fact, prompted the experiment.

Detecting a Signal in Noise: Further Considerations

In discussing the masking of tonal signals, L. A. Jeffress, H. C. Blodgett, T. T. Sandel, and C. L. Wood [J. Acoust. Soc. Am. 28, 416 (1956)] likened the auditory mechanism for detection to the kind of equipment a physicist might use in attempting to perform the same task. Both involve the use of a narrow filter followed by a detector. The earlier paper described the detector as needing “to be sensitive only to changes in over-all level and to have a certain time characteristic.” The present paper is concerned with properties of the detector, especially its time characteristics for both monaural and binaural stimulation. Some of the problems of signal detection will be illustrated by oscillographic pictures of narrow-band noise and signals for both the homophasic and the antiphasic conditions. (This work was supported by the Bureau of Ships.)

Some Determinants of Interaural Phase Effects

It has recently been shown that the binaural threshold for a pure tone presented against a background of noise depends upon the interaural phase differences of the tone and the noise. For example, the threshold of a low frequency tone, in phase at the two ears, is low when it is heard against noise that is out of phase at the two ears, but is high when it is heard against noise that is in phase. The masked threshold of a 250-cycle tone presented against a background of intense noise is in fact about 15 db lower under antiphasic conditions than it is under homophasic conditions. (Homophasic refers to the condition in which the tone and the noise have the same interaural phase difference. Antiphasic refers to the condition in which either tone or noise, but not both, is out of phase at the two ears.) When the tone is masked by another tone, fairly close in frequency but not so close that beats are detected, the dependence of masking on interaural phase does not appear. The present experiment attempts to determine what characteristics masking sound (“masker”) must have if it is to produce interaural phase effects. A 250-cycle tone was presented against four different kinds of background. The masking sounds were (1) pure tones, (2) regular pulses (125 p.p.s.), (3) random pulses (average 125 p.p.s.) and (4) random noise. Combinations of filters permitted the presentation of these maskers in frequency bands that were varied in width and center frequency. The results indicate that the threshold for the tone in the presence of a regular, periodic masker does not depend to any significant degree upon interaural phase relations. A necessary characteristic of the masker that produces interaural phase effects is randomness or irregularity. Both masking and the interaural phase effects increase as the frequency band of a random masker approaches the frequency of the tone. Band width alone does not contribute to the interaural phase effects except as widening a band tends to bring one cut-off nearer the frequency of the tone. Wide and narrow band maskers, both of which contain the frequency of the tone, produce the same effects.

Some Determinants of Interaural Phase Effects

Recently the binaural masked threshold has been shown to depend upon the phase angle between the two ears for both the masked signal and the masking signal. These interaural phase effects are particularly clear for pure tones of fairly low frequency (100–800 c.p.s.) that are masked by white noise. It has been shown that a pure tone, in-phase at the two ears, that is presented against a background of white noise is more easily heard when the noise is out-of-phase than when the noise is in-phase at the ears. The converse is also true, namely, that if the tone is out-of-phase it is more easily heard when the noise is in-phase. The masked threshold of a 250-cycle tone presented against a background of noise (100 db SPL in a 7000-cycle low pass band) is approximately 15 db lower under antiphasic (tone-in, noise-out or tone-out, noise-in) than under homophasic (tone-in, noise-in or tone-out, noise-out) conditions. When the same tone is masked by another tone whose frequency is fairly close but not close enough to produce beats, no such differences appear. These two masking signals represent the extremes of a continuum of complexity along which the necessary characteristics of the adequate stimulus for these differences should appear. The present experiment constitutes an attempt to find these characteristics. A pure tone at 250 c.p.s. was used as the signal to be masked throughout the experiment. Four different kinds of masking signals were used: pure tones; ‘regular’ pulses (125 p.p.s.), ‘random’ pulses (average 125 p.p.s.) and random noise. The signals were presented in frequency-bands which were varied in respect of band-width and center frequency. The results indicate that a regular, periodic masking signal will not produce these interaural phase effects. A necessary condition is randomness or irregularity with respect to time but not necessarily with respect to amplitude. Frequency spectrum does not enter on any such all-or-none basis but rather contributes to the magnitude of the response. The nearer a frequency-band is to the frequency of the masked tone, the greater are the differences between the homophasic and antiphasic conditions. There is no substantial difference between the interaural phase effects produced by a narrow versus a wide band, provided that both bands contain the frequency of the masked tone.