Psychophysical and Physiological Evidence for a Precedence Effect in the Median Sagittal Plane

Abstract

A listener in a room is exposed to multiple versions of any acoustical event, coming from many different directions in space. The precedence effect is thought to discount the reflected sounds in the computation of location, so that a listener perceives the source near its true location. According to most auditory theories, the precedence effect is mediated by binaural differences. This report presents evidence that the precedence effect operates in the median sagittal plane, where binaural differences are virtually absent and where spectral cues provide information regarding the location of sounds. Parallel studies were conducted in psychophysics by measuring human listeners' performance, and in neurophysiology by measuring responses of single neurons in the inferior colliculus of cats. In both experiments the precedence effect was found to operate similarly in the azimuthal and sagittal planes. It is concluded that precedence is mediated by binaurally based and spectrally based localization cues in the azimuthal and sagittal planes, respectively. Thus, models that attribute the precedence effect entirely to processes that involve binaural differences are no longer viable.

Full text

77:2223-2226, 1997. J Neurophysiol
Ruth Y. Litovsky, Brad Rakerd, Tom C. T. Yin and William M. Hartmann
You might find this additional information useful...
18 articles, 3 of which you can access free at: This article cites http://jn.physiology.org/cgi/content/full/77/4/2223#BIBL
4 other HighWire hosted articles: This article has been cited by
[PDF] [Full Text] [Abstract] , September 1, 1998; 80 (3): 1285-1301. J Neurophysiol
R. Y. Litovsky and T. C. T. Yin I. Correlates of Psychophysics
Physiological Studies of the Precedence Effect in the Inferior Colliculus of the Cat.
[PDF] [Full Text] [Abstract] , September 1, 2001; 86 (3): 1333-1350. J Neurophysiol
B. J. Mickey and J. C. Middlebrooks
Localization
Responses of Auditory Cortical Neurons to Pairs of Sounds: Correlates of Fusion and
[PDF] [Full Text] [Abstract] , October 1, 2003; 90 (4): 2149-2162. J Neurophysiol
D. J. Tollin and T. C.T. Yin Psychophysical Investigation of an Auditory Spatial Illusion in Cats: The Precedence Effect
[PDF] [Abstract] , November 1, 2008; 70 (8): 1558-1570. Atten Percept Psychophys
L. D. SANDERS, A. S. JOH, R. E. KEEN and R. L. FREYMAN One sound or two? Object-related negativity indexes echo perception
on the following topics: http://highwire.stanford.edu/lists/artbytopic.dtlcan be found at Medline items on this article's topics
Physiology .. Humans
Physiology .. Cats
Veterinary Science .. Inferior Colliculus
including high-resolution figures, can be found at: Updated information and services http://jn.physiology.org/cgi/content/full/77/4/2223
can be found at: Journal of Neurophysiologyabout Additional material and information http://www.the-aps.org/publications/jn
This information is current as of January 28, 2009 .
http://www.the-aps.org/.Physiological Society. ISSN: 0022-3077, ESSN: 1522-1598. Visit our website at
by the American Physiological Society, 9650 Rockville Pike, Bethesda MD 20814-3991. Copyright © 2005 by the American
publishes original articles on the function of the nervous system. It is published 12 times a year (monthly)Journal of Neurophysiology
on January 28, 2009 jn.physiology.orgDownloaded from

RAPID COMMUNICATION
Psychophysical and Physiological Evidence for a Precedence Effect in
the Median Sagittal Plane
RUTH Y. LITOVSKY,
1
BRAD RAKERD,
2
TOM C. T. YIN,
1
AND WILLIAM M. HARTMANN
3
1
Department of Neurophysiology, University of Wisconsin-Madison, Madison, Wisconsin 53706; and
2
Department of
Audiology and Speech Sciences and
3
Department of Physics and Astronomy, Michigan State University, East Lansing,
Michigan 48824
Litovsky, Ruth Y., Brad Rakerd, Tom C. T. Yin, and William
is symmetrical with respect to the two ears. Although micro-
M. Hartmann. Psychophysical and physiological evidence for a
phone measurements normally find some binaural differ-
precedence effect in the median sagittal plane. J. Neurophysiol.
ences for sources in the sagittal plane (Searle et al. 1975;
77: 2223–2226, 1997. A listener in a room is exposed to multiple
Wightman and Kistler 1989), the preponderance of evidence
versions of any acoustical event, coming from many different direc-
shows that these differences are not reliable enough to serve
tions in space. The precedence effect is thought to discount the
as localization cues (Asano et al. 1990; Hebrank and Wright
reflected sounds in the computation of location, so that a listener
1974; Middlebrooks and Green 1991). Instead, sources in
perceives the source near its true location. According to most audi-
the median sagittal plane are localized on the basis of spectral
tory theories, the precedence effect is mediated by binaural differ-
shape cues, peaks and valleys introduced by direction-depen-
ences. This report presents evidence that the precedence effect
operates in the median sagittal plane, where binaural differences
dent filtering performed by the external ears, head, and torso
are virtually absent and where spectral cues provide information
(Blauert 1983; Gardner and Gardner 1973; Roffler and But-
regarding the location of sounds. Parallel studies were conducted
ler 1968).
in psychophysics by measuring human listeners’ performance, and
The present study is a search for a precedence effect in the
in neurophysiology by measuring responses of single neurons in
median sagittal plane, a precedence effect that is mediated by
the inferior colliculus of cats. In both experiments the precedence
spectral shape cues. Early evidence that such an effect likely
effect was found to operate similarly in the azimuthal and sagittal
exists was found in a front-back competition experiment by
planes. It is concluded that precedence is mediated by binaurally
Blauert (1971). The present search was conducted on two
based and spectrally based localization cues in the azimuthal and
levels: one psychophysical in humans, the other physiologi-
sagittal planes, respectively. Thus,models that attribute the prece-
cal in cats.
dence effect entirely to processes that involve binaural differences
are no longer viable.
METHODS
Psychophysics
INTRODUCTION
The psychophysical experiments were performed in a 32-m
3
A room plays havoc with sound. The waves emitted by
anechoic room with the use of loudspeakers to simulate direct
a source are reflected and rereflected many times by the
sounds and reflections. There were five matched speakers: directly
room surfaces. Therefore a listener in a room is exposed to
in front at 0
7
, behind, 90
7
to the left, 90
7
to the right, and overhead.
multiple versions of any acoustical event, coming from many
On each experimental trial there were eight pairs of leading and
lagging clicks (each 0.025 ms in duration) repeated every 110 ms.
different directions in space. The auditory system can cope
Such a click train allows the precedence effect to build to a maxi-
with this sonic clutter because of the precedence effect, a
mum (Freyman et al. 1991). Trials were presented in blocks con-
remarkable neural process that fuses the direct sound and
sisting of trials in the azimuthal plane (left, front, and right sources )
its reflections into a single image (Haas 1951; Wallach et
or the sagittal plane (front, overhead, and behind sources). All
al. 1949). The precedence effect also discounts the reflected
permutations of leading source location, lagging source location,
sound in the computation of location so that a listener per-
and eight values of interclick delay (ICD ) ranging from 0 to 10
ceives the source near its true location. The standard theoreti-
ms were tested. The experiments measured localization in a compe-
cal model for this effect is an extension of the binaural model
tition experiment, so that after the stimulus was presented the sub-
for localization, a neural coincidence detector that operates
ject had to decide which of the three loudspeakers in the plane
on the difference in arrival time of signals at left and right
was closest to the location of the sound image. Eight subjects each
completed a total of 10 blocks for each plane, with blocks for the
ears (Jeffress 1948). The extension postulates an inhibitory
different planes randomly interspersed.
response generated by the leading sound (Franssen 1963;
Lindemann 1986). Physiological evidence for the coinci-
dence detector has been found in the medial superior olive Physiology
(Goldberg and Brown 1969), and psychophysical evidence
Physiological experiments paralleled the psychophysical ones by
for the extension to precedence has been found in headphone
using click sources in a free-field, anechoic chamber and comparing
experiments (Zurek 1980).
azimuthal and sagittal plane responses. Extracellular recordings
The binaural difference model successfully describes
were made in 38 neurons (characteristic frequencies ranging from
many aspects of localization, but it does not account for the
500 to 24,000 Hz) in the central nucleus of the inferior colliculus
(ICC) of barbiturate-anesthetized cats. The animal’s head was in
localization of sources in the median sagittal plane, which
22230022-3077/97 $5.00 Copyright
q
1997 The American Physiological Society
J561-6RC/ 9k0f$$ap24 08-27-97 15:17:05 neupal LP-Neurophys
on January 28, 2009 jn.physiology.orgDownloaded from

R. Y. LITOVSKY, B. RAKERD, T.C.T. YIN, AND W. M. HARTMANN2224
the center of semicircular arrays (radius of 1.2 m) of loudspeakers
Physiology
positioned along the azimuthal and sagittal planes at 15
7
intervals.
Positive angles refer to sounds in the contralateral hemifield in
In the physiological experiments for each neuron, we first
azimuth and above the interaural line in elevation. The precedence
measured the receptive field properties with the use of single
effect was probed by delivering clicks from two different speakers
clicks (0.1 ms in duration) at a level
Ç
10–15 dB above
with varying ICDs, with 50 trials for every condition. We measured
threshold (Fig. 2A). To compare the degree of suppression
precedence by the degree of suppression of the response to the
for stimuli on the azimuthal and sagittal planes, we always
lagging click as a function of the presence of a leading click at
chose the speaker directly in front, which lies at the intersec-
different ICDs.
tion of the two planes, to be the lagging source; and to
control for possible influences of response rate, we chose
RESULTS
the locations of the two leading sources such that the re-
sponses to a single click were approximately equal. In Fig.
Psychophysics 2 the leading sources were placed at
/
90
7
azimuth and
/
75
7
elevation. The dot rasters in Fig. 2Bshow that, for leadingFigure 1 shows the percentage of responses that matched
the leading click location at each ICD. If the precedence clicks in the azimuthal plane, at long ICDs (
ú
40 ms), there
is a response at a latency of
Ç
16 ms to the leading clickeffect is operating, this percentage will be high. Open sym-
bols show the results in the azimuthal plane, where binaural and a later response at about the same latency following the
lagging click. The response to the lagging click graduallydifference cues are present. These results are in good agree-
ment with previous studies of precedence (Yost and Soder- diminishes as the ICD is shortened and disappears for ICDs
õ
31 ms. Figure 2Cshows nearly identical behavior whenquist 1984; Zurek 1980) . When ICDs were
õ
1 ms, the
precedence effectwas incomplete;leading andlagging clicks the leading click was in the sagittal plane. Figure 2Dshows
little difference in the suppression of the response to theboth affected the perceived location to some degree, an effect
that has been called summing localization (Warncke 1941). lagging click in the two different planes when plotted as
normalized recovery curves.At ICDs of 1.0 and 2.0 ms, precedence with binaural differ-
ences was maximal, although still not entirely complete; To quantify the extent of suppression, from the recovery
curves we measured the ICD at which the lagging responsesubjects chose the leading click location on as many as 95%
of all trials. As the ICD increased to 5.0 ms, the lagging was suppressed by 50%. The similarity in the values of half-
maximal suppression shown in Fig. 2Dis typical of thatclick began to be audible, and it was chosen as the location
of the sound source on some trials, indicating that the prece- observed in most cells in thepopulation, although the overall
shapes of the recovery curves in Fig. 2Dare more similardence effect became weaker. The filled symbols in Fig. 1
show the results for the sagittal plane experiment. They show than usually seen. Figure 3 shows a scatter plot of the ICDs
for half-maximal suppression along the azimuthal and sagit-that a spectrally mediated precedence effect exists, and that
its dependence on ICD is comparable with that of the binau- tal planes for 38 cells. There are two striking features of the
data shown in Fig. 3. First, there is considerable variabilityral differenceprecedence effect. The only appreciable differ-
ence is that the spectrally mediated effect was somewhat in the degree of suppression between different cells in the
ICC, ranging from
Ç
0to
ú
100 ms. Such variability hasweaker at its maximum.
FIG
. 1. Mean responses for 8 subjects. For both source
planes, the plot shows the percentage of trials in which the
sound image appeared at the position of the leading source.
High percentages indicate a strong precedence effect. Similar
functions of the interclick delay (ICD) occur for both planes,
although the precedence effect is stronger when binaural dif-
ferences are present (open symbols).
J561-6RC/ 9k0f$$ap24 08-27-97 15:17:05 neupal LP-Neurophys
on January 28, 2009 jn.physiology.orgDownloaded from

PRECEDENCE IN THE SAGITTAL PLANE 2225
FIG
. 2. Physiological evidence for precedence in the azimuthal and sagittal planes for 1 neuron. A: azimuthal and sagittal
rate functions. At each location, 50 repetitions of clicks were presented with a period of 300 ms. The number of spikes per
stimulus is plotted against azimuthal and sagittal locations. Arrows: leading source location for Band C. B and C: dot rasters
showing the responses of the same neuron at ICDs ranging from 1 ms ( bottom) to 101 ms (top ) along the azimuthal (left)
and sagittal (right) planes. D: recovery functions for the lagging responses shown in Band C, normalized by the response
to the same stimulus in absence of the leading stimulus.
previously beenseen in theICC under other conditions (Car- ally. Although the dominance of the leading source in local-
ney and Yin 1989; Yin 1994). Second, for each cell there ization is maximal
Ç
2 ms, suppression of the lag as an
is a strong correlation (r
Å
0.8) between its suppressive independent auditory event can extend to 50 ms, as in echo
effect in the two planes. Thus for any given cell in the ICC suppression in concert halls (Kuttruff 1979). Localization
the degrees of suppression of the lagging click to a leading dominance of the lead was the only behavioral measure
click in the sagittal and azimuthal planes are comparable. probed by the psychophysical experiments above.
In summary, the experiments reported here show that the
precedence effect operates similarly in the azimuthal and
DISCUSSION
sagittal planes. Because the degree of suppression varies
The variability in suppression delays mirrors the range of with changes in relative location of the leading and lagging
sources in both planes (Litovsky and Yin 1994), precedencedelays at which the precedence effect is observed behavior-
J561-6RC/ 9k0f$$ap24 08-27-97 15:17:05 neupal LP-Neurophys
on January 28, 2009 jn.physiology.orgDownloaded from

R. Y. LITOVSKY, B. RAKERD, T.C.T. YIN, AND W. M. HARTMANN2226
FIG
. 3. Correlation of 50% maximal response along the
azimuthal vs. sagittal planes. Each asterisk represents the
values for 1 neuron.
superior olivary complex to dichotic tonal stimuli: some physiological
is mediated by binaurally based and spectrally based local-
mechanisms of sound localization. J. Neurophysiol. 32: 613 –636, 1969.
ization cues, in the azimuthal and sagittal planes, respec-
F
RANSSEN
,N.V.Stereophony. Eindhoven, The Netherlands: Philips Techni-
tively. Therefore models that attribute the precedence effect
cal Library, 1963. (English translation 1964.)
entirely to processes that involve binaural differences are no
F
REYMAN
, R. L., C
LIFTON
, R. K.,
AND
L
ITOVSKY
, R. Y. Dynamic processes
longer viable. However, precedence is not identical in all
in the precedence effect. J. Acoust. Soc. Am. 90: 874–884, 1991.
planes; the psychophysical localization experiments show
H
AAS
, H. On the influence of a single echo on the intelligibility of speech.
that it is stronger in the azimuthal plane. This difference
Acustica 1: 49–58, 1951.
might be related to the fact that neurons in the ICC show
H
EBRANK
,J.H.
AND
W
RIGHT
, D. Spectral cues used in the localization of
stronger firing rate variation with source position in the azi-
sound sources on the median plane. J. Acoust. Soc. Am. 56: 1829 –1834,
1974.
muthal plane compared with the sagittal plane. Despite the
J
EFFRESS
, L. A. A place theory of sound localization. J. Comp. Physiol.
quantitative differences, precedence appears to operate simi-
Psychol. 61: 468–486, 1948.
larly whether the localization cues are binaural or spectral.
K
UTTRUFF
,H.Room Acoustics (2nd ed.) . London: Applied Science Publish-
ers, 1979.
L
INDEMANN
, W. Extension of a binaural cross-correlation model by contra-
This research was supported by National Institute of Deafness and Other lateral inhibition. I. Simulation of lateralization for stationary signals. J.
Communications Disorders Grant DC-00181 to Michigan State University Acoust. Soc. Am. 80: 1608–1622, 1986.
and Grant DC-00116 to the University of Wisconsin. L
ITOVSKY
,R.Y.
AND
Y
IN
, T.C.T. Physiological correlates of the precedence
Present addresses of R. Litovsky: Dept. of Biomedical Engineering, Bos- effect: free-field recordings in the inferior colliculus of the cat (Abstract).
ton University, 44 Cummington St., Boston, 02215; and Eaton Peabody Assoc. Res. Otolaryngol. 85, 1994.
Laboratory, Massachusetts Eye and Ear Infirmary, Boston, MA 02215. M
IDDLEBROOKS
,J.C.
AND
G
REEN
, D. M. Sound localization by human
Address for reprint requests: R. Litovsky, Dept. of Biomedical Engi- listeners. Annu. Rev. Psychol. 42: 135–159, 1991.
neering, Boston University, 44 Cummington St., Boston, MA 02215. R
OFFLER
,S.K.
AND
B
UTLER
, R. A. Factors that influence the localization
of sounds in the equatorial plane. J. Acoust. Soc. Am. 43: 1255 –1259,
Received 16 July 1996; accepted in final form 5 December 1996. 1968.
S
EARLE
, C. L., B
RAIDA
, L. D., C
UDDY
, D. R.,
AND
D
AVIS
, M. F. Binaural
pinna disparity: another auditory localization cue. J. Acoust. Soc. Am.
REFERENCES 57: 448–455, 1975.
W
ALLACH
, H., N
EWMAN
, E. B.,
AND
R
OSENZWEIG
, M. R. The precedence
A
SANO
, F., S
UZUKI
, Y.,
AND
S
ONE
, T. Role of spectral cues in median plane effect in sound localization. Am. J. Psychol. 57: 315–336, 1949.
localization. J. Acoust. Soc. Am. 88: 159–168, 1990. W
ARNCKE
, H. The fundamentals of room-related stereophonic reproduction
B
LAUERT
, J. Localization and the law of the first wavefront in the median in sound films. Akust. Z. 6: 174–188, 1941.
plane. J. Acoust. Soc. Am. 50: 466–470, 1971. W
IGHTMAN
,F.L.
AND
K
ISTLER
, D. J. Headphone simulation of free-field
B
LAUERT
,J.Spatial Hearing. Cambridge, MA: MIT Press, 1983. listening. II. Psychophysical validation. J. Acoust. Soc. Am. 85: 868–
C
ARNEY
,L.H.
AND
Y
IN
, T.C.T. Responses of low-frequency cells in the 878, 1989.
inferior colliculus to interaural time differences of clicks: excitatory and Y
IN
, T.C.T. Physiological correlates of the precedence effect and summing
inhibitory components. J. Neurophysiol. 62: 144–161, 1989. localization. J. Neurosci. 14: 5170–5186, 1994.
G
ARDNER
,M.B.
AND
G
ARDNER
, R. S. Problem of localization in the medial Y
OST
,W.A.
AND
S
ODERQUIST
, D. R. The precedence effect: revisited. J.
plane: effect of pinnae cavity occlusion. J. Acoust. Soc. Am. 53: 400 – Acoust. Soc. Am. 76: 1377 –1383, 1984.
408, 1973. Z
UREK
, P. M. The precedence effect and its possible role in the avoidance
of interaural ambiguities. J. Acoust. Soc. Am. 67: 952–964, 1980.G
OLDBERG
,J.M.
AND
B
ROWN
, P. B. Responses of binaural neurons of dog
J561-6RC/ 9k0f$$ap24 08-27-97 15:17:05 neupal LP-Neurophys
on January 28, 2009 jn.physiology.orgDownloaded from