p a p e r s
p r e s e n t a t i o n s

Vassilakis, P.N. (2004b).  Dichotic beats: Perception of sound interference versus detection of sound-source motion.  Proceedings of the 8th ICMPC (International Conference on Music Perception and Cognition).  R. Ashley, S. Lipscomb, editors (CD-ROM).  Evanston, IL: Northwestern University.


It has been argued that beating-like sensations, arising when sines with slightly different frequencies are presented dichotically, are manifestations of sound interference occurring at neural or cognitive levels. The present study examines the relationship between dichotic beats and the perception of sound-source motion, as well as the perceptual similarities/differences between signals with interaural intensity modulations (IIMs) and interaural phase modulations (IPMs – equivalent to interaural frequency differences that give rise to dichotic beats). In a categorization experiment, subjects were asked to make similarity judgments among stimuli that included pure tones, amplitude modulated tones, tones with IIMs, and tones with IPMs. The results indicate that, when two sines with slightly different frequencies are presented dichotically, the phase-based sound-localization perceptual mechanism may come into play, with the constantly shifting phase between the dichotically presented sines (i.e. IPM) effecting a constantly shifting localization of the two-component complex stimulus. Depending on the average and difference frequencies of the dichotic components, this shift may be interpreted as sound-source rotation and/or timbral fluctuation, often confused with loudness fluctuation. The findings do not support the claim that sound interference products can arise from the processing of a combined binaural channel. They are consistent with previous sound localization studies and suggest that, as is the case with IIMs, the auditory system has difficulty resolving fast IPMs. The importance of interaural phase- rather than interaural time-differences to sound localization is discussed, addressing implications to sound-image creation in headphone listening.

Background and Aims: Beating-like sensations, usually associated with the phenomenon of interference, have been reported even in studies where stimuli with slightly different frequencies are presented dichotically through headphones and are not allowed to physically interfere. Previous explanations have implicated cross-hearing, ‘interference’ at neural levels, or cognitive processes. It is argued that interference-like perceptions, reported in response to dichotic stimuli with interaural frequency differences (IFDs), are not manifestations of sound interference but are linked to sound localization cues and the detection of sound-source motion.
Method: 20 subjects were presented with a series of categorization tasks. In each task they were given 3 model stimuli [a diotic steady tone, a diotic sinusoidally amplitude-modulated (AM) tone, and a dichotic ‘rotating tone’ (tone with interaural intensity modulations, IIMs)], under which they were asked to group 4 experiment stimuli [the 3 model stimuli and one dichotic tone with IFDs of 1-20Hz, equivalent to interaural phase modulation (IPM) rates of 1-20Hz]. The modulation parameters of the AM and IIM signals were determined during a preliminary experiment matching AM, IIM, and IPM signals at all IPM rates. Subjects listened to the stimuli through headphones, repeating the categorization task for all IPM rates and at 4 stimulus frequencies representing a low- and a high-frequency region.
Results: Tones with IIMs were identified correctly at low interaural modulation rates. As the modulation rate increased, they were increasingly confused with the AM tones, especially at high-frequencies. Categorization of the IPM tones depended on the average frequencies of the stimuli. At high average frequencies, IPM tones were categorized as steady tones. At low average frequencies and for low IPM rates they were grouped under the IIM tones. As the IPM rate increased, responses where increasingly shared between IIM and AM tones. Perception of sound-source motion persisted over a wider range of interaural modulation rates for pulse-like versus sinusoidal modulations.
Conclusions: In dichotic explorations of beats, the constantly shifting phase between the components presented in the two ears results in a constantly shifting localization of the two-component stimulus. At low frequencies (<~500Hz), IPMs are treated perceptually as IIMs. For low modulation rates (<~10Hz) they both support the perception of sound-source rotation. As the rate of interaural modulation increases, the sound-source rotation sensation gradually turns into timbral fluctuation, occasionally perceived as loudness fluctuation. At high frequencies (>~1500Hz), dichotic tones with IPMs are treated perceptually as pure tones. Dichotic tones with IIMs continue to be perceived as rotating tones and, as the rate of interaural modulation rises, they are increasingly confused with AM tones.
The results are consistent with previous sound localization studies. They suggest that dichotic beats of mistuned unisons are a between-channel phase effect and a manifestation of a sound-source motion-detection mechanism, with the binaural interaction observed corresponding to interpretation of sound localization cues rather than interference. Subjects find it equally difficult to resolve IPMs and IIMs, and both IPMs and IIMs are resolved over a wider range of modulation rates for stimuli with pulse-like rather than sinusoidal modulations. The temporal integration and the resulting ‘smoothening’ assumed to be at the base of binaural ‘sluggishness’ may therefore occur together for IPMs and IIMs at low stimulus frequencies. The frequency-dependent roles of IPMs to the perception of sound-source rotation and of interaural time differences to sound localization, suggest that sound localization at low frequencies may be aided by interaural phase rather than interaural time differences.