Some Measurements of Interaural Time Difference Thresholds

R. G. Klumpp and H. R. Eady

Citation: The Journal of the Acoustical Society of America 28, 859 (1956); doi: 10.1121/1.1908493
View online: http://dx.doi.org/10.1121/1.1908493
View Table of Contents: http://asa.scitation.org/toc/jas/28/5
Published by the Acoustical Society of America

Articles you may be interested in

The dominant role of low-frequency interaural time differences in sound localization
The Journal of the Acoustical Society of America 91, 1648 (1998); 10.1121/1.402445

Model for the interaural time differences in the azimuthal plane

The Journal of the Acoustical Society of America 62, 157 (1998); 10.1121/1.381498

The precedence effect

The Journal of the Acoustical Society of America 106, 1633 (1999); 10.1121/1.427914

On the Minimum Audible Angle

The Journal of the Acoustical Society of America 30, 237 (2005); 10.1121/1.1909553

The effect of head-induced interaural time and level differences on speech intelligibility in noise
The Journal of the Acoustical Society of America 83, 1508 (1998); 10.1121/1.395906

Lateralization of High-Frequency Tones

The Journal of the Acoustical Society of America 32, 132 (2005); 10.1121/1.1907864
THE JOURNAL OF THE ACOUSTICAL SOCIETY OF AMERICA VOLUME 28, NUMBER 5 SEPTEMBER, 1956

Some Measurements of Interaural Time Difference Thresholds

R. G. KLU'MP AND H. R.
U.S. Navy ElectronicsLaboratory,San Diego52, California
(ReceivedAugust8, 1955)

Thresholdsfor the detectionof interaural time differencewere determinedby ten listeners(1) for band-
limited randomnoise(150-1700cps), (2) for a 1000-cpstone,and (3) for a 1-millisecond click.The average
interauraltime differences
corresponding to 75% correctdetectionin the symmetricaltwo-alternativetests
were (1) 9 microseconds,(2) ll microseconds, and (3) 28 microseconds.
Rangesof individualthresholdsand
grouppsychometricfunctionsare presented.

TABLE I. Interaural time difference thresholds in microseconds

Ninteraural
1953preliminary measurements
time difference were
thresholdofmadeofthe
a group of
for varioussignalsas read from curveslike thoseshownin Fig. 1.

listenersfor a variety of sounds.A larger programof Threshold Number Judgments

measurements,then planned, was never carried out. Signal
in
microseconds
of
listeners
per point
per listener
Sincethe resultsof the preliminary measurements may
Noise Broad band s 10 10 160
be of interest to others, they are presentedhere in 150-1700cpsb 9 10 80
abbreviated form. 425-600 cpsb 14 9 160
A two-channel circuit containing electrical delay 410440 cpsb 19 8 80
2400-3400 cpsb 44 10 80
linesand terminatingin earphoneswasusedto control, 3056-3344cpsb 62 10 80
independently,the time and intensity relationshipsof
Clicks (1 Single 28 10 120
signalsdeliveredto the earsof a listener. In one channel millisecond
the signalcouldbe delayedfrom 0 to !000 microseconds duration) Repeated, 30 clicks
in a 2 second burst 11 13 160
in 1-microsecond stepsand in the other channelfrom
0 to 900 microseconds in 18-microsecond steps. Tones 90 cps 75 10 10
In the method used, the listener heard a pair of 125 cps 56 9 80
250 cps 27 9 80
sounds,a standardand a variable, and judged which 500cps 17 9 80
was further to the left. The standard was presented 1000cps 11 9 80
with zero interchannel time difference and the variable 1300 cps 24 10 10
1500 cps ..- 10 10
with an interchanneltime differenceleadingin the left 1800cps ... 10 10
ear. The order of presentation of the two sounds 3200 cps .-- 10 80
comprisinga pair was randomized.Each signalwas at
maximumintensity for approximately1.4 secondsand Random noise limited by the frequency response of the PDR-8 ear-
phones used.
was turned on and off over 0.3 second. A 2-second b Half-power points.

interval of silence separated the two. Interchannel

intensitydifferencewaszero for both sounds,hencethe Thresholdsfor noisesignalsappear to be much the
only clueuponwhichthe listenercouldbasehis discrimi- sameas thresholdsfor pure toneswithin the samepass
nation was interchannel time difference. band as the noise. Thus, the threshold of 14 micro-
Results for three of the sounds tested are shown in seconds for a 425-600 cpsnoiseband is the sameas the
Fig. 1. The soliddots representresultsobtainedwith a estimatedthresholdfor a 600-cpstoneand the threshold
band of noise generated by passing random noise
through a half-octave filter set to 150-1700 cps band
I00,
pass. The crossesare data obtained with a 1000-cps
tone. The circles are for a click of 1-millisecond duration. 9O

The interaural time difference threshold (75%

so
o
correctpoint) of ten listenersfor the noiseband is 9
microseconds. Individual thresholdsranged from 5 to
7o
18 microseconds. For the 1000 cps tone, the group
thresholdis 11 microseconds with a rangefrom 7 to 23
microseconds. For the click, the group thresholdis zso
28 microseconds with a range from 19 to 46 micro-
$00 I0 s)O $0 40 50 60 70 80 90
seconds. INTERGHANNEL TIME OIFFERENGE OF VARIABLE
IN MIGROSEGONOS
Table I lists interaural time difference thresholds for
other signalsalong with the number of listenersand Fro. 1. The percent judgments correct as a function of the
interchannel time difference of the variable sound. The number
the number of judgmentsfor each of the five points of judgmentsper point per listenerand the number of listeners
determiningthe thresholdcurve. used are listed in Table I.

859
860 R. G. KLUMPP AND H. R. EADY

of 19 microseconds for a 410-440 cpsnoiseband is the shownin Fig. 1, the thresholdfor a 150-1700cpsnoise
sameas the estimatedthresholdfor a 440-cps tone. band under this condition is 9 microseconds. When the
This relationship wouldnot applyfor frequencies above standard has an interchannel time difference of 430
about 1300 cps since listeners were unable to detect microseconds
leadingin the right ear, the thresholdis
changes in interchannel time difference for tones of 29 microseconds. With the standard at 790 micro-
1500, 1800, and 3200 cps. However, the thresholdof seconds,the thresholdis 50 microseconds.
Thus, the
62 microseconds for the 3056-3344cpsnoiseband which thresholdfor this noise band increasesby about !
hadno audiblecomponents below2000cpssuggests that microsecondfor every20 microsecondsof displacement
sensitivity to time differenceis not restricted to low of the standard.
frequencieswhen the signalis a noise. All of the listenersusedhad normalhearingin both
The measurements listed in Table I were made with ears. Pure tones were presented at 65 db and noise
the standard at zero interchannel time difference. As signalsat from 60 to 80 db over-allre 0.0002dyne/cma.

THE JOURNAL OF THE ACOUSTICAL SOCIETY OF AMERICA VOLUME 28, NUMBER 5 SEPTEMBER, 1956

Just Noticeable Differences in Dichotic Phase*

J. Zws.oci Am) R. S. Fm.r)A$
Psycho-Acoustic
Laboratory,
HarvardUniversity,Cambridge,
Massachusetts
(ReceivedFebruary3, 1956)

Thejustnoticeable difference
in dichotic
phase,asa functionofsensationlevelandoffrequency,hasbeen
determined on a numberof listenerswith normalhearing.The testtonesweretransmitted by earphones,
andthephasedifference between theearswasvariedby meansof an electronic phaseshifter.Thepsycho-
physicalmethod usedcombined pairedcomparisons andforcedchoice.Thefirsttonepulseof eachpairpre-
sentedwaskeptat a constant phasedifferenceat whichthesubjectlocalizedthesoundsource asequidistant
fromhisears.The dichotic phasedifferenceof thesecondpulsewasvariedirregularly("randomly").The
resultsshowthat the sensitivity
to dichoticphasedifference
is highest(2 of phase)at mediumsensation
levels,andthat thejndincreases withpositiveacceleration
asthesoundfrequency increases.Around1300
cps the jnd becomesso great that it cannot be measured.The dichotictime differencecalculatedfrom the
measuredjnd in phasehas a minimumnear 800 cps.

HEpsychophysical
descriptions literature
of the contains
measuresof justanumber of hasAlmost
noticeable
every psychophysical experiment,however,
led to a differentvaluefor the attribute measured,
differences.
In auditionalone,measures
of jnd havebeen with the result that the reader of the literature is
made on all three basic attributes of acoustic vibration offereda wide choiceof jnds for any givenattribute.
--intensity, frequency,and phase. The jnd seemsto be particularlydependenton the
500cps,90db psychophysicalmethod used. This dependencyis
SENSATION LEVEL
emphasized in papersby W. A. Rosenblithand K. N.
CONSTANT STIMULI
Stevens and by Pollack? Rosenblith and Stevens
mentionfour classes of jnd for frequency,the AX, the
'x _ 7.5 ol- o ABX, the quantal, and the modulationind. For the
' 80[-.- UP
AND
DOWN jnd in intensityPollackhasfounda "floatingstandard,"
x 6
"single standard," "single comparison standard,"
"roving standard,"and "roving comparisonstandard"
ind.
No classificationhas beenfound for the jnd in phase
becauseof the paucity of reference.And, of the few
-- I I I I I I I I I I [ I I t t t t t scattered referencesin the literature, the majority
LEFT PHASEDIFFERENCE
IN DEGREES GHT do not concernpure phase,but are expressedas time
differences.
m. . umultive distribution curves. 1osed circles indicate
the percentageo tone pulsesjudged to be to the right o the The experiments describedhereinmustbe considered
standard pulse; the crossesindicate the percentage of pulses aspreliminaryrather than final.We did not try to make
uded to be to the left. he open circlesjoined by dashedline
correspondto vlues clculted [rom te up-nd-downmethod. our measurements very precise,in the expectationthat
* his work ws crried out under ontrct N5ori-76 between they wouldnot representthe last wordin any case,and
rvrd niversity nd the Oce of Nv] Research,. S.
(Proect NR142-201, Report PNR-82). Reproductionor W. A. Rosenblithand K. N. Stevens,J. Acoust.Soc.Am. 25,
purposeo the . S. Governmentis permitted. 980 (1953).
On leave rom te versity o sseusetts. 'I. Pollack,J. Acoust.Soc.Am. 26, 1056(1954).