Page 24 - Summer 2008

tones presented in rapid succession to the two ears to perceive melodies; the perceptual fusion of simultaneous input to the two ears, particularly to determine pitch; the issue of binaural rivalry; and binaural beats. All these have been examined in recent times, and we briefly examine those recent experiments that are particularly related to the earlier work.
Largely for technological reasons, research on the local- ization of sounds in the early and middle parts of the twenti- eth century focused on very simple sound signals, such as pure tones and noise. Much of this work was inspired by Lord Raleigh’s duplex theory of binaural perception.56 Sound local- ization by the binaural system (according to Raleigh’s duplex theory) is based largely on two types of cue— interaural time and intensity differences—the latter particularly at low fre- quencies. The use of these cues has been confirmed and stud- ied extensively (see Stern et al58 for a review). For single sounds presented in isolation, the binaural system has been shown capable of utilizing interaural time-differences of around 10 microseconds, and interaural intensity-differences
59,60
analysis, considerable interest also developed in the involve- ment of spectral cues for sound localization, particularly
61
58
binaural integration was addressed to the nature of the sounds that were perceived, independently of how they were localized. The early thought experiment proposed by Wells10 concerning integration of melodic patterns with tones presented in alter- nation to the two ears was very similar to one carried out by Deutsch,64 who was, at the time, unaware of this earlier work. In this experiment, continuously repeating melodies were pre- sented at a rate of roughly 6 per second, and listeners were asked to identify them. When the tones were presented dioti- cally, the identification rate was very high. However, when the tones were switched haphazardly between the ears, listeners were unable to integrate them into a single perceptual stream, and so to identify the melody. This finding was in accordance
10
had surmised. When a low drone was presented contralater- al to the ear receiving each melody component, identification again rose to high levels. From this and other control condi- tions it was concluded that when each ear received input sep- arately, tones were organized by spatial location; however, when presentation of the contralateral drone enabled an alternative organization by pitch proximity, listeners instead integrated the tones from the two ears in accordance with
of around 1 dB.
With further improvements in instrumentation and
Attention here has focused on spectral differences produced by the head, and by the outer and middle ears. The insertion of probe micro- phones in the ear enabled the determination of the transfer function from the sound source to the eardrum. Using this technique, it has been found, in particular, that interaural intensity differences for different frequency bands provide important cues to the localization of sounds coming from
along the vertical and front-back planes.
62,63
developments, sophisticated models of binaural interaction
different directions.
In parallel with these experimental
for sound localization were developed.
As described above, much of the early work concerning
with the prediction made by Wells.
However, the reason for this difficulty was not as Wells
Nicholas Wade combines art, science and history in his “perceptual portraits.” His training is in visual psy- chology and he developed an interest in the interaction between visual art and visual science from his analysis of “op art.” Having taught himself how to produce works in this genre, he combines drawn designs with photograph- ic images, often producing works that are at the limits of visibility. His art work has been published in his books: The Art and Science of Visual Illusions (Routledge and Kegan Paul, 1982), Visual Allusions: Pictures of Perception (Erlbaum, 1990), Psychologists in Word and Image (MIT Press, 1995) and Circles: Science, Sense and Symbol (Dundee University Press, 2007). He is currently working on Galileo and the Senses with Marco Piccolino, which will be published by Oxford University Press.
The faces shown in perceptual portraits are not always easy to discern—the viewer needs to apply the power of perception in order to extract the facial features from the design which carries the portraits. They generally consist of two elements—the portrait and some appropriate motif. The nature of the latter depends upon the endeavours for which the portrayed person was known. Thus, Chladni (Fig. 1) is embedded in his acoustic figures; Wheatstone (Fig. 4) is in his stereoscope; Weber (Fig. 5) is shown three times but two portraits are near to threshold in curves that correspond to psychometric functions (the colors repre- sent those he examined for binocular color rivalry); Fechner (Fig. 6) is hidden in the area of uncertainty of psy- chometric functions; and Thompson (Fig. 8) is combined in complementary colors in the normal head orientation and left-right reversed, so that the eyes are reversed as in pseudoscopic vision. Similar principles have been applied to the portraits of the authors (see page 27). That of Wade is based on binocular interaction as the centers of the intersecting circles that carry the dimly defined face are located at each of the eyes. The second is of Deutsch. It alludes to binaural integration with two waveforms approaching the sides of the head and combining in the midline, which corresponds to the midline of the embed- ded portrait.
Some examples of perceptual portraits can be found at: http://www.perceptionweb.com/wade/
and in many of the editorial essays in Perception; see: http://www.dundee.ac.uk/psychology/njwade/
Nicholas J. Wade n.j.wade@dundee.ac.uk
this principle. Thus the difficulty in melody identification found in the primary experiment was not due to an inability to integrate information across ears at fast rates, but rather to a grouping decision made by the auditory system that could be overridden by other cues (see also Bregman65).
Work on integrating streams of speech that were switched between the ears can be interpreted in the same light. Cherry and Taylor66 presented a single stream of speech
Binaural Hearing and the Stethophone 23