Page 53 - Summer 2015

ing” procedures to measure Fechnerian-type psychophysi- cal relationships between subjective/perceptual attributes of sound and physical measures. Stevens power law is a fre- quently used psychophysical relationship.
Smitty was famous in an entirely different arena from psy- chology and psychophysics. He is often touted as the father of “short skis.” Smitty was avid skier at a time when skis were expected to be as tall as one could reach. Smitty felt this was “nonsense” and that shorter skis (far shorter) would offer much more movability with only a small loss in speed. He and others with a passion for short skis convinced the ski world of their merit, and the tall skis of the past are just that, in the past.
By the 1940s, it was well understood that ITDs and ILDs controlled sound source localization in the horizontal plane (see earlier discussion of Rayleigh). By presenting sounds over headphones, ITDs and ILDs could be systematically varied and independently controlled. The perception of sounds presented over headphones is not always the same as when the same sounds are presented in an open field. Thus the study of ITD and ILD cues using headphone-delivered sounds is referred to as “lateralization” as opposed to “lo- calization” when the sounds are presented by loudspeakers.
In 1948 in the same issue of the Journal of the Acoustical Soci- ety of America (JASA), Ira Hirsh and J. C. R. Licklider (“Lick” was known for his early work on the internet and personal computing; see Waldrop, 2001; Walden, 2014; Table 1) each published an article showing that when a tonal (Hirsh, 1948)
Figure 8. Depiction of masking level difference (MLD) conditions. When the signal (top waveform) and masker (bottom waveform) are the same at both ears, detection is difficult. Deleting the signal from one ear (leading to an interaural time difference [ITD] and an inter- aural level difference [ILD]) renders the signal easier to detect. From Green and Yost (1975).
Figure 9. “Jeffress Coincidence Network” with inputs from the left and right ears converging in a central neural area of bipolar cells. From Jeffress (1948).
or speech (Licklider, 1948) signal was presented with a dif- ferent configuration of ITDs/ILDs than those of the noise masker, the threshold for detecting the signal was lower (sometimes a whopping 15 dB lower) than when the signal and masker had the same ITD/ILD configuration. This im- provement in threshold was called the “masking level dif- ference (MLD; Figure 8).” Due to the connection between lateralization/localization and the MLD and the large size of the MLD, MLD-like studies have been a dominant psycho- acoustic topic since the 1950s.
In 1948, Lloyd Jeffress (Table 1) proposed that ITDs could be produced by a neural coincidence network that could be modeled as a cross-correlator (Figure 9). The “Jeffress model” has been widely used and forms the basis for under- standing how barn owls locate small rodents at night using only sound. Many years later, Nat Durlach (1963; Table 1) developed a model of the MLD, the equalization-cancella- tion (EC) model, which assumes that the binaural auditory system first attempts to equalize the sounds at each ear and then interaurally cancels the equalized sounds. Studies of binaural processing and testing various models occupied many pages in JASA from 1948 until today and form the ba- sis of current models of Spatial Hearing (Colburn, 1973; see also Blauert, 1997; Table 1).
Summer 2015 | Acoustics Today | 51