The Octave Illusion Revisited Again
University of California, San Diego
The octave illusion (D. Deutsch, 1974) occurs when 2 tones separated by an octave are alternated
repeatedly, such that when the right ear receives the high tone, the left ear receives the low tone, and vice
versa. Most subjects in the original study reported hearing a single tone that alternated from ear to ear,
whose pitch also alternated from octave to octave, and D. Deutsch (1975a) proposed an explanation in
terms of separate what and where auditory pathways. C. D. Chambers, J. B. Mattingley, and S. A. Moss
(2002) argued that the perceived pitch difference generally corresponds more to a semitone and proposed
an alternative explanation in terms of diplacusis. This article argues that Chambers et al. used problematic
procedures and reports a new experiment on the octave illusion. The findings confirm that an octave
difference is generally perceived, and they agree with the model of Deutsch (1975a) but are at variance
with the diplacusis hypothesis.
The octave illusion, which was originally described by Deutsch
(1974), is a paradoxical auditory phenomenon that is characterized
by large individual differences in perception. The pattern that was
first used to create this illusion consisted of two tones that were
spaced an octave apart and that were repeatedly presented in
alternation. The identical sequence was presented via headphones
to both ears simultaneously; however, when the right ear received
the high tone, the left ear received the low tone, and vice versa.
This pattern gave rise to a number of different illusory percepts,
the most common one (termed octave) being of a single tone that
alternated from ear to ear, whose pitch also alternated from one
octave to the other in synchrony with the localization shift.
Deutsch (1975a) proposed a model to account for the octave
percept, based on a hypothesized separation between what and
where decision mechanisms in the auditory system. The model,
hereafter referred to as the two-channel model, assumes that (a) to
produce the perceived pitches, the frequencies arriving at one ear
are perceived, while those arriving at the other ear are suppressed
from conscious perception and that (b) each perceived tone is
localized at the ear receiving the higher frequency signal, regard-
less of whether a pitch corresponding to the higher or lower
frequency is in fact perceived. The model therefore assumes that
the octave illusion results from illusory conjunctions of pitch and
Chambers, Mattingley, and Moss (2002) argued from a series of
experiments that the phenomenology of the octave illusion differs
from that originally described by Deutsch (1974). More specifi-
cally, they asserted that, on listening to the octave illusion, the
perceived difference between the alternating tones generally cor-
responds more to a semitone than to an octave. On this basis, the
authors hypothesized that the tones at the two ears fuse harmoni-
cally to produce a pitch that corresponds to the low tone,1and that
the slight pitch difference between the alternating tones that is
perceived is the result of diplacusis.2They also reported that some
of their subjects lateralized each tone to the ear receiving the
higher frequency and that some lateralized each tone to the ear
receiving the lower frequency, though they did not offer an expla-
nation for the lateralization patterns they obtained.
In this article, I first review early findings concerning the
phenomenology of the octave illusion and describe the two-
channel model that was proposed to explain the octave percept.
There follows a critique of the study by Chambers et al. (2002),
which questions the validity of their observations. Because their
hypothesis could hold only if these observations were valid, this
hypothesis is also challenged. Finally, a new experiment is re-
ported in which the phenomenology of the octave illusion is
documented more explicitly than in previous studies. The findings
from this new experiment confirm that an octave difference be-
tween the alternating tones is generally perceived, and they are in
accordance with the two-channel model but cannot be explained
on the diplacusis hypothesis.
Previous Findings Concerning the Phenomenology of the
Octave Illusion and the Two-Channel Model
In the original experiment of Deutsch (1974), subjects were
presented with dichotic sequences consisting of 250-ms tones. As
shown in Figure 1, the tones alternated between 400 Hz and 800
Hz, such that when the right ear received 400 Hz, the left ear
1When two sine-wave tones that stand in octave relation are presented
simultaneously, they fuse perceptually so that a single tone is heard, whose
pitch corresponds to the low tone (the fundamental frequency). This effect
of harmonic fusion has also been found to occur when the tones are
presented dichotically (Houtsma & Goldstein, 1972).
2The term diplacusis refers to a slight difference in perceived pitch (of
a small fraction of a semitone) that can occur when the same tone is
presented to the left and right ears (van den Brink, 1975a, 1975b). This
effect is usually attributed to the structural characteristics of the ear.
I thank Trevor Henthorn for his valuable assistance in setting up the
instrumentation for the experiment.
Correspondence concerning this article should be addressed to Diana
Deutsch, Department of Psychology, University of California, San Diego,
La Jolla, CA 92093. E-mail: firstname.lastname@example.org
Journal of Experimental Psychology:
Human Perception and Performance
2004, Vol. 30, No. 2, 355–364
Copyright 2004 by the American Psychological Association
received 800 Hz, and vice versa. The tones were sine waves, at
equal amplitude, with no gaps between tones. To minimize tran-
sients, there were no amplitude drops at the frequency transitions,
and phase continuity was preserved at the transitions.
Eighty-six subjects (53 right-handers and 33 left-handers), all
naive concerning the octave illusion, were presented with a 20-s
segment of the illusion and asked to report what they heard. The
positions of the earphones were then reversed, and the procedure
was repeated. A number of different illusory percepts were ob-
tained, and these were divided into three categories. The first,
termed octave, consisted of a single tone that alternated from ear
to ear, whose pitch also alternated from one octave to the other in
synchrony with the localization shift. For most subjects who ob-
tained this percept, when the positions of the earphones were
reversed, the apparent locations of the high and low tones did not
reverse with them.3The octave category of percept was described
in 58% of right-handers and 52% of left-handers. Furthermore,
those right-handers who obtained this percept showed a strong
tendency to hear the high tone on the right and the low tone on the
left; however, the left-handers as a group did not preferentially
localize the high and low tones either way.
The second category of percept, termed single pitch, consisted
of a single tone that alternated from ear to ear, whose pitch either
did not change or changed only slightly with the shift in the tone’s
perceived location. This percept was described in 25% of right-
handers and 9% of left-handers. The third category, termed com-
plex, comprised a number of different complex percepts, often
involving at least three different pitches. This category of percept
was obtained in 17% of right-handers and 39% of left-handers. The
two handedness groups differed significantly, both in terms of type
of percept obtained (with a higher proportion of left-handers
reporting complex percepts) and also in terms of patterns of later-
alization for the octave percept (with a higher proportion of right-
handers reporting the high tone on the right).4
Deutsch (1983b) further studied handedness correlates with
perception of the octave illusion in 250 subjects. The tendency to
perceive the high tone on the right was found to be higher among
right-handers than mixed-handers, and it was higher among mixed-
handers than left-handers.5Furthermore, for all three handedness
groups, the tendency to hear the high tone on the right was lower
among those with left- or mixed-handed parents or siblings than
among those with only right-handed parents and siblings. This
pattern of results is in accordance with the neuropsychological
literature relating patterns of cerebral dominance to handedness
and familial handedness background (Herron, 1980) and leads to
the conjecture that perception of the octave illusion might serve as
a reflection of the direction and degree of cerebral dominance in
Other work on the octave illusion was directed toward under-
standing the basis of the type of percept termed octave. Deutsch
(1975a) hypothesized that this percept results from a dissociation
between the what and where pathways in the auditory system. The
proposed two-channel model is shown in Figure 2. To produce the
perceived pitches, the listener follows the frequencies that are
presented to the dominant ear and suppresses from conscious
perception those that are presented to the nondominant ear. How-
ever, the listener lateralizes each perceived tone to the ear receiv-
ing the higher frequency signal, regardless of whether he or she
perceives a tone corresponding to the higher frequency or the
lower one. Because the pitch and lateralization decision mecha-
nisms here use different rules, illusory conjunctions of pitch and
Take, as an example, a listener whose pitch perceptions corre-
spond to the frequencies presented to his or her right ear. When the
high tone is presented to the right ear and the low tone is presented
to the left ear, this listener perceives the high tone, because this
tone is presented to the right ear. This listener also lateralizes the
tone to his or her right ear, because this ear is receiving the higher
frequency signal. However, when the low tone is presented to the
right ear and the high tone to the left ear, the listener hears the low
tone, because this is presented to the right ear, but the listener
lateralizes the tone to his or her left ear instead, because this ear is
receiving the higher frequency signal. So the listener hears the
entire sequence as a high tone to the right alternating with a low
tone to the left. However, now take a listener whose pitch percep-
tions correspond to the frequencies presented to the left ear, and
hold the lateralization rule constant. This listener hears the same
3The earphone positions were reversed, rather than reversing the signals
delivered to the earphones, in order to control for possible differences in
4In the experiment described by Deutsch (1974), the pitch differences
perceived on listening to the octave illusion were documented by verbal
reports, which were backed up either by musical notation or by marking the
perceived pitch difference on a scale from zero to over an octave. Because
of space limitations, these methods were not described in this article.
5Handedness and familial handedness background were evaluated using
the questionnaire and procedure of Varney and Benton (1975).
(1974) describing the octave illusion, together with the percept most
commonly obtained. Black boxes indicate tones at 800 Hz, and white boxes
indicate tones at 400 Hz. From “An Auditory Illusion,” by D. Deutsch,
1974 (September 27), Nature, 251, p. 307. Copyright 1974 by Macmillan
Publishers Ltd. Adapted with permission.
The stimulus pattern used in the original experiment of Deutsch
stimulus pattern as a high tone to the left alternating with a low
tone to the right.6
To test this hypothesis, Deutsch and Roll (1976) presented 44
right-handers with repeating sequences consisting of 400–800-Hz
dichotic tone pairs. One ear received a repeating pattern consisting
of three high (800-Hz) tones alternating with two low (400-Hz)
tones, while simultaneously the other ear received a repeating
pattern consisting of three low (400-Hz) tones alternating with two
high (800-Hz) tones. The tones were 250 ms in duration and were
separated by 250-ms pauses. The subjects made two judgments: (a)
how many high tones they heard in sequence and how many low
tones they heard in sequence—so indicating which ear they fol-
lowed for pitch—and (b) how many tones they heard in sequence
in the right ear and how many in the left ear—so indicating to
which ear each tone was lateralized. The results were as predicted
from the two-channel model. Concerning the pitch component,
most subjects reported hearing the patterns of high and low tones
that were presented to their right ear rather than their left. How-
ever, each tone was lateralized to the ear that received the higher
frequency signal, regardless of whether a pitch corresponding to
the higher or lower frequency was perceived.
Further experiments (Deutsch, 1978, 1980, 1981, 1988) used the
two-alternative forced-choice (2AFC) method to investigate per-
ception of the octave illusion under parametric manipulation. Here,
subjects were selected who showed strong and stable octave per-
cepts. To explore the ear dominance component, I presented sub-
jects with segments of the illusion and asked them to report on
each trial whether they heard a pattern that began with the high
tone and ended with the low tone (i.e., a high–low–high–low
pattern) or a pattern that began with the low tone and ended with
the high tone (i.e., a low–high–low–high pattern). In this way, the
subjects indicated which ear they were following for pitch. The
relative amplitudes of the tones at the two ears were varied, and the
percentages of judgments that corresponded to the frequencies
presented to the nondominant ear were plotted as a function of
these amplitude relationships. The strength of the ear dominance
effect was then measured by the size of amplitude difference
between the tones at the two ears required to counteract it. The
effect was found to be strong for sequences in which the two ears
received the same frequencies in succession (i.e., when both the
400-Hz and the 800-Hz tones were presented in immediate suc-
cession to the left and right ears, as in the original octave illusion
pattern), but it was weaker or absent for sequences in which this
pattern of relationship did not hold (Deutsch, 1980). The effect
was also found to be weaker when the interonset interval between
successive tones was lengthened to 3,000 ms, regardless of
whether this was achieved by inserting silent gaps between the
tones or by increasing the durations of the tones themselves
To explore the lateralization component, I again presented sub-
jects with segments of the illusion but now asked them to report on
each trial whether they heard a pattern that began at the left ear and
ended at the right ear (i.e., a left–right–left–right pattern) or a
pattern that began at the right ear and ended at the left ear (i.e., a
right–left–right–left pattern). In this way, the subjects indicated
whether the tones were lateralized to the ear receiving the higher
frequency or the lower one. The relative amplitudes of the higher
and lower tones were varied, and the percentages of judgments that
corresponded to the ear receiving the lower frequency were plotted
as a function of these amplitude relationships. It was found that the
subjects lateralized the tones to the ear receiving the higher fre-
quency until the amplitude of the lower tones exceeded those of
the higher ones by a critical level. This effect occurred with tones
presented in rapid repetitive sequence (as in the original octave
illusion pattern) but was significantly weaker when only two
dichotic tone pairs were presented (Deutsch, 1978) or when more
complex pitch configurations were used (Deutsch, 1988).
Zwicker (1984) explored the phenomenology of the octave
illusion in several experiments. Concerning the issue of perceived
pitch differences between the alternating tones, the author re-
6Deutsch (1981) elaborated on this model to provide a more concrete
explanation of the octave illusion in terms of its neurophysiological un-
derpinnings. The elaborated model also accommodates findings from later
parametric studies of the illusion, including those showing a dependence on
the durations between onsets of successive tones. Because of space limi-
tations, this model is not described here.
and the other determining perceived location, can combine to produce the octave illusion. Black boxes indicate
tones at 800 Hz, and white boxes indicate tones at 400 Hz. R ? right; L ? left. From “Auditory Illusions,
Handedness, and the Spatial Environment,” by D. Deutsch, 1983, Journal of the Audio Engineering Society, 31,
p. 608. Copyright 1983 by the Audio Engineering Society. Adapted with permission.
Illustration showing how the outputs of two decision mechanisms, one determining perceived pitch
ported, from the judgments of 15 subjects, that “mostly octaves,
but also often smaller intervals were perceived [italics added]” (p.
129). He also confirmed that, for 400-Hz and 800-Hz signals, there
was a strong tendency to lateralize the tones to the ear receiving
the higher frequency. In a further experiment, he presented 8
subjects with octave illusion patterns at tone durations (and, so,
interonset intervals) ranging from 0.01 s to 2.00 s. He wrote,
The observers’ certainty in perceiving Deutsch’s illusion . . . showed
a clear maximum with tone durations around 200 ms; with decreasing
tone durations other acoustic illusions appear, while with durations
greater than about 1 s, the presentation can be perceived correctly. (p.
Figure 3 plots the percentages of different perceptions of the
illusion pattern at the different tone durations (and, so, interonset
intervals) in Zwicker’s study, and it can be seen that, at interonset
intervals of 2,000 ms, over 80% of reports were of no illusion.
Study of Chambers et al. (2002)
Chambers et al. (2002) challenged the report of Deutsch (1974)
concerning the octave illusion. From their experiments, they con-
cluded that the perceived pitch difference between the alternating
tones generally corresponds more to a semitone than to an octave.
They further concluded that listeners do not uniformly lateralize
each tone toward the ear receiving the higher frequency. On the
basis of these claims, the authors proposed an alternative expla-
nation of the octave illusion. This explanation assumes that
1. listeners perceptually fuse the dichotically presented high
and low tones, so as to perceive a pitch that corresponds
roughly to the fundamental frequency;
2. the slight pitch difference that is heard between the
alternating tones is the result of diplacusis; and
3. the perceived tones are sometimes lateralized to the ear
receiving the higher frequency and sometimes to the ear
receiving the lower frequency.
The present section presents a critique of the study by Chambers
et al. (2002) and questions the validity of their observations. Given
that their model could hold only if these observations were valid,
their theoretical proposal is also challenged. The study consisted of
four experiments, and these are examined below.
Experiment 1, titled “Subjective Report,” was intended to be
preliminary; as Chambers et al. (2002) wrote, “Subjective reports
were purely qualitative and were not statistically analyzed [italics
added]” (p. 1292). Fifteen subjects participated; 3 of these were the
authors, and no information was given concerning whether or not
the remaining subjects were naive concerning the octave illusion.
The subjects first listened extensively to single dichotic tone pairs
at 400 Hz and 800 Hz, at several tone durations, which ranged
from 200 to 800 ms. The dichotic 400–800-Hz tones were then
presented in alternating sequence. The tone durations again ranged
from 200 to 800 ms. Furthermore, on half the trials, the tones were
separated by pauses that were equal in duration to the tones
themselves, so the interonset intervals between successive tones
varied from 200 ms to 1,600 ms.
The issue of tone duration.
Chambers et al. (2002) stated that
the subjects’ judgments exhibited no dependence on tone duration
nor on the presence of regular silent intervals, and they claimed
that this was in accordance with findings obtained by others.
However, the authors also stated that the subjects’ reports were not
statistically analyzed, and they presented no data from patterns at
specific tone durations; neither did they present any other evidence
to support their assertion that judgments were indeed independent
of tone duration. Further, the authors were incorrect in asserting
that their observations were in accordance with previous ones. As
outlined above, Deutsch (1981) compared interonset intervals of
250 ms with those of 3,000 ms and obtained a highly significant
effect of interonset interval (see also Deutsch, 1983a). This study
used subjects who had been selected for obtaining a strong octave
percept in the first place, which leads to the surmise that unselected
subjects might show an even stronger effect of interonset interval.
Such a result was obtained in the study by Zwicker (1984) refer-
enced above. As shown in Figure 3, from Zwicker’s data, one
would expect that at interonset intervals of 800 ms, no illusion
would be perceived roughly 30% of the time, and one would
expect that at interonset intervals of 1,600 ms, no illusion would be
perceived roughly 80% of the time. Yet, these values of interonset
interval were among those used by Chambers et al. (2002).
The issue of lateralization.
In Experiment 1, Chambers et al.
(2002) used the following method to evaluate how each tone was
illusion, as a function of tone duration and thus of interonset interval. The
data were averaged across 8 subjects. Percept 1: Two tones of identical
pitch alternating from ear to ear. Percept 2: A higher tone in one ear
alternating with a lower tone in the other ear (i.e., the octave percept).
Percept 3: A higher tone alternating from ear to ear, together with a lower
tone alternating from ear to ear (i.e., no illusion). Percept 4: None of the
above. From “Experimente zur dichotischen Oktav-Tauschung,” by T.
Zwicker, 1984, Acustica, 55, p. 135. Copyright 1984 by S. Hirzel Verlag,
Stuttgart, Germany. Adapted with permission.
Percentage occurrence of different percepts of the octave
lateralized. The subjects were presented with alternating 400–
800-Hz sequences at the various interonset intervals described
above and were instructed to indicate their judgments by tapping
(e.g., “Tap in time with the higher pitch” or “Tap in time with the
left tone”). The experimenter then evaluated, by visual observation
of the subjects’ tappings, to which ear the higher and the lower
tones were lateralized. The authors reported, on the basis of these
observations, that 9 subjects lateralized the tones toward the ear
receiving the lower frequency, whereas 6 subjects lateralized the
tones toward the ear receiving the higher frequency.
However, the authors presented no actual data to support this
assertion. They did not state that the experimenter monitored the
signals that were presented to the subjects, and it is unclear how,
without such monitoring, he could determine how the subjects’
tappings corresponded to their perceptions. Even assuming that the
experimenter had monitored the sound signals, no evidence was
provided that he could reliably synchronize his visual perceptions
of the subjects’ tappings with these signals, nor that the subjects
were able to synchronize their tappings with these signals. This
problem of validity is particularly severe for the fast rates of
presentation needed to produce the octave illusion; yet, at slow
rates of presentation, the illusion becomes degraded and may even
disappear (Figure 3). It is important to note that these informal
observations concerning lateralization in Experiment 1 were not
confirmed elsewhere in the article by Chambers et al. (2002). In
contrast, a number of other experiments, which were reviewed
above, have indicated that, on listening to the octave illusion, there
is a strong tendency for subjects to lateralize each perceived tone
toward the ear receiving the higher frequency signal (Deutsch,
1978, 1988; Deutsch & Roll, 1976; Zwicker, 1984).
The issue of the size of the perceived pitch difference.
evaluate the size of the perceived pitch difference between the
alternating tones, Chambers et al. (2002) presented the subjects,
for comparison, with sequences in which tones alternated between
400 Hz and 800 Hz (an octave) and between 400 Hz and 424 Hz
(a semitone). The authors reported that 2 of the subjects perceived
a pitch difference of an octave, 8 (2 of whom were authors)
perceived a pitch difference of a semitone, 4 (1 of whom was an
author) perceived a pitch difference of between an octave and a
semitone (this difference not being further specified), and 2 sub-
jects perceived no pitch difference. However, several points should
here be made.
1. The subjects’ judgments were based on sequences in
which the interonset intervals varied substantially, in-
cluding some in the range where Zwicker (1984) had
reported that sometimes no illusion was obtained. Com-
parison cannot, therefore, be made directly between these
results and those of Deutsch (1974).
2. The subjects had earlier been given extensive experience
with dichotic 400–800-Hz chords and patterns at these
different tone durations, and this prior experience may
have influenced their judgments.
3. At least 3 of the subjects (i.e., the authors) had prior
knowledge of the illusion, and this could have influenced
4. Twenty-five percent of right-handers in the original
large-scale experiment of Deutsch (1974) reported little
or no pitch difference between the alternating tones (i.e.,
their percepts fell into the single pitch category). In
Experiment 1 of Chambers et al. (2002), the authors
reported on the responses of only 12 subjects (excluding
themselves), so that sampling bias could have contributed
to their results.
However, as the authors pointed out, Deutsch (1974) did not
publish the means by which the subjects’ reports of the octave
illusion were obtained (see Footnote 4). This is rectified in a new
experiment, which is described later in this article.
In Experiment 2, Chambers et al. (2002) did not investigate the
octave illusion itself, but rather a different effect, and they related
their results to the informal observations reported in Experiment 1.
The experiment used 8 subjects who had all participated in Ex-
periment 1, including 2 authors (the 2 subjects who had reported
hearing an octave difference between the tones in Experiment 1
were not tested). The subjects were presented with tones at 400 Hz
or at 800 Hz to one ear, and they were asked to match the tone that
they heard to a tone of variable pitch that was presented to the
other ear, thereby obtaining a measure of diplacusis. The subjects’
matches varied between no pitch difference to a difference of less
than half a semitone, with the majority of matches being in the
lower part of this range. None of the matches involved a pitch
difference that approached a semitone.
On the basis of these findings, Chambers et al. (2002) claimed
that pitch differences obtained on hearing the octave illusion are a
reflection of diplacusis. However, the very slight pitch differences
they found in Experiment 2 were inconsistent with this interpre-
tation, given their reports in Experiment 1 of an octave difference
by 2 subjects and of differences that were greater than a semitone
but smaller than an octave by 4 additional subjects. So, even
setting aside the problems of interpretation outlined above, and the
previous reports of an octave difference between the alternating
tones by other researchers (Deutsch, 1974; Zwicker, 1984),
diplacusis could only account for some of the reports of the octave
illusion obtained by Chambers et al. in their own Experiment 1.
In Experiment 3, subjects were presented with dichotic se-
quences of various types and were asked on each trial to report
which ear was receiving the higher pitch. The subjects were not
asked to report their actual percepts but, rather, to infer what
signal configurations were being presented. Eight subjects were
tested. These had all participated in Experiment 1 and so had
received extensive experience with octave illusion patterns of
differing durations, some of which may well have been perceived
veridically (Zwicker, 1984). Chambers et al. (2002) wrote, “The
most surprising result from this experiment was the capacity
listeners demonstrated to correctly segregate the octave illusion
sequence by ear, despite reporting a standard single-image per-
cept” (p. 1297). However, this result is unsurprising given the
subjects’ prior experience with variants of the illusion in Experi-
ment 1, some of which would have enabled them to infer what
signals were being presented to each ear.
Seven subjects participated in Experiment 4. These had all
participated in Experiment 1, though their percepts of the octave
illusion in this experiment were not given. The subjects were
presented with dichotic 400–800-Hz sequences, in which tones at
400 Hz, 800 Hz, or 2,000 Hz were embedded as deviants. The
deviant tones were presented diotically (i.e., simultaneously to
both ears) but offset in time so that they were perceptually dis-
placed to the side of the midline. Averaged across subjects, reac-
tion times were shorter for detecting 800-Hz than 400-Hz deviants.
Chambers et al. (2002) interpreted this finding as indicating that,
on hearing the standard octave illusion, listeners would perceive
both of the alternating tones as closer to 400 Hz than to 800 Hz.
However, the finding by Deutsch (1980) that perception of the
octave illusion is sensitive to sequential context leads to the
possibility that the diotically presented tones may have affected
perception of the illusion here also. Furthermore, the data were
averaged across all 7 subjects, so that the inclusion of even 1 or 2
subjects whose percepts fell into the single pitch category would
have skewed it in the direction reported by Chambers et al.
In sum, taking together the four experiments by Chambers et al.
(2002), the following can be noted:
1. Chambers et al.’s (2002) conclusions relied heavily on
the informal observations from Experiment 1, which the
authors themselves stated were “purely qualitative and
were not statistically analyzed [italics added]” (p. 1292).
These findings were based on perceptions of sequences
of tones with widely differing interonset intervals, in-
cluding some in the range where Zwicker (1984) had
reported that sometimes no illusion was obtained. Other
aspects of the procedures used in this study (such as the
tapping procedure for establishing lateralization patterns)
were problematic, and many of the observations were at
variance with those of Deutsch (1974, 1978, 1980, 1981,
1988), Deutsch and Roll (1976), and Zwicker (1984).
2. Chambers et al.’s (2002) interpretation of the pitch dif-
ferences heard in the octave illusion in terms of diplacu-
sis cannot explain the pitch differences of an octave that
were reported by others (Deutsch, 1974; Zwicker,
1984)—or even a sizable proportion of the pitch differ-
ences that were reported by the authors themselves in
their Experiment 1.
However, Chambers et al. (2002) have challenged the claim
made by Deutsch (1974) that the majority of subjects, on listening
to the octave illusion, perceive a pitch difference of an octave
between the alternating tones. Instead, they have argued that the
most common percept of this illusion involves a very small pitch
difference (i.e., of a semitone or less). Their argument was based
on the informal observations of a small number of subjects, some
of whom were the authors themselves, and all of whom had
experienced intensive prior training with dichotic chords at differ-
ent durations before listening to the illusion. However, given that
the authors have raised this issue, an experiment was carried out to
examine the size of the perceived pitch difference between the
alternating tones on listening to the octave illusion, using the
stimulus parameters that had originally been used by Deutsch
(1974). This new experiment used a procedure that enabled more
explicit conclusions to be drawn than those from earlier
In this experiment, musically trained subjects were asked to
listen to the octave illusion pattern and to write down in musical
notation what they heard. The subjects were furnished with the
note name of the low tone in the pattern (G4) and were told that
this was one of the tones they would hear. They then used relative
pitch to notate the other tones that they perceived. If the octave
illusion simply reflected diplacusis, one should expect that, on
hearing this pattern, the subjects would notate only a small pitch
difference between the alternating tones. To control for the possi-
bility that the subjects might mistakenly attribute an octave differ-
ence between tones of the same pitch that were presented to the left
and right ears, two further patterns were presented. The first
consisted of the high tone of the octave illusion pattern alone (G5)
alternating from ear to ear, and the second consisted of the low
tone of the octave illusion pattern alone (G4) alternating from ear
male, ages 18–32 years) participated in the experiment. They had all
received at least 4 years of formal musical training and could read and
write in simple musical notation. On the basis of the handedness question-
naire of Varney and Benton (1975), they comprised 8 right-handers (3 with
left- or mixed-handed relatives), 3 mixed-handers, and 1 left-hander. All
subjects were naive concerning the octave illusion.
Apparatus and stimuli.
Three sequences of sine-wave tones were cre-
ated. The first, the octave illusion pattern, was as in the experiment of
Deutsch (1974). This constituted a sequence of tones that alternated be-
tween 400 Hz and 800 Hz (corresponding to approximately G4 and G5 in
the musical scale). The tones were of equal amplitude and were 250 ms in
duration. To minimize transients, there were no amplitude drops between
tones, and the frequency transitions occurred at zero crossing.7The iden-
tical sequence was presented to both ears simultaneously; however, when
the right ear received the high tone, the left ear received the low tone, and
vice versa. The second, the alternating high tone pattern, consisted of tones
at 800 Hz (corresponding approximately to G5) that were presented in
alternation at the two ears. All tones were at equal amplitude and were 250
ms in duration. The third, the alternating low tone pattern, was identical to
the second, except that the tones were at 400 Hz (corresponding approxi-
mately to G4).
Tones were generated on a NeXTStation Turbo (NeXT Computers, Inc.,
Redwood City, CA) using the cmusic sound synthesis system (F. R. M.
Moore, 1982). The signals were transferred to a Macintosh G4 computer,
passed through a mixer (Mackie CR1604; LOUD Technologies, Inc.,
Woodinville, WA) and then through an amplifier (NAD 304; NAD Elec-
Twelve individuals with normal hearing (5 male and 7 fe-
7The tones were generated in phase at the two channels.
tronics, Sharon, MA), and were presented to subjects via headphones
(Grason-Stadler TDH-49, calibrated and matched; Grason-Stadler, Inc.,
Madison, WI) at an amplitude of 70-dB SPL. The subject was seated in
front of a Keystation 61 synthesizer keyboard (M-Audio, Irwindale, CA)
that was interfaced with the computer, so that the subject was able to play
on the keyboard and hear the output through earphones while at the same
time listening to the test patterns.
All subjects were tested individually, and they listened to
the patterns through earphones. They were told that on each trial, they
would hear a repeating sequence of tones and that they should notate both
the sequence of pitches they heard and the perceived ear of input for each
tone. They were given no initial practice sequences.
The subjects first listened to the octave illusion pattern for as long as
they wished. They were informed that one of the tones approximated G4
but were given no further information. They were enabled to confirm their
perceptions by matching the tones in the pattern with tones they played on
the synthesizer keyboard. When they were certain of their judgments, they
notated the sequence of pitches they perceived, together with the perceived
locations of the tones. Following this, the subjects were asked to place their
earphones in reverse position (see Footnote 3), to repeat the procedure, and
to notate again the sequence of pitches they perceived, together with the
perceived locations of the tones. Next, the alternating high tone pattern was
presented, and the same procedure was followed. Then, the alternating low
tone pattern was presented, and the same procedure was followed. Finally,
the octave illusion pattern was again presented, and the subjects were asked
to report the pitch differences that they heard.
Figure 4 presents, as examples, the notations of 3 of the subjects.
Subjects SY and RR notated the octave illusion pattern as a tone
corresponding to G5 on the right, alternating with a tone corre-
sponding to G4 on the left, with earphones placed both ways. In
contrast, Subject JP notated the same pattern as a tone correspond-
ing to G5 on the left, alternating with a tone corresponding to G4
on the right, with earphones placed both ways. When presented
with patterns consisting of single tones (i.e., G5 alternating from
ear to ear, and G4 alternating from ear to ear), all these subjects
notated the tones correctly.
The data from all 12 subjects are summarized in Table 1. On
listening to the octave illusion pattern, 7 of the 12 subjects notated
The notations for Track 1 (a) are of the octave illusion pattern and for Track 1 (b) are of the octave illusion
pattern with earphone positions reversed. The notations for Track 2 are of a single tone at G5 (high tone), which
alternated from ear to ear, and the notations for Track 3 are of a single tone at G4 (low tone), which alternated
from ear to ear.
Percepts of the different patterns in the experiment, notated by 3 of the subjects (SY, RR, and JP).
the standard octave percept described by Deutsch (1974), that is, a
single tone that alternated from ear to ear, that simultaneously
alternated between G4 and G5, with earphones placed both ways.
Of these, 4 subjects heard the high tone on the right and the low
tone on the left, with earphones placed both ways; 2 heard the high
tone on the left and the low tone on the right with earphones placed
both ways; and 1 heard the high tone on the right and the low tone
on the left on one presentation, with the opposite lateralization
pattern on the other. On relistening to the octave illusion pattern at
the end of the session, all these subjects again reported hearing
tones an octave apart that alternated from ear to ear. Four addi-
tional subjects notated complex percepts that changed with con-
tinued listening, all of which involved G4 and G5, and so involved
an octave difference between the tones. (One of these subjects
notated G4 alternating from ear to ear, interspersed with other
notations involving both G4 and G5.) The percepts of these 4
subjects therefore fell into the complex category described in
Deutsch (1974). The final subject notated the same pitch (G4)
alternating from ear to ear, with the tone in the right ear having a
sharper timbre. The percept of this subject therefore fell into the
single pitch category described by Deutsch (1974).
On listening to the alternating high tone pattern, 11 subjects
correctly notated G5 alternating from ear to ear, whereas 1 subject
notated a semitone difference between the tones at the two ears
(i.e., G5 alternating with G#5). On listening to the alternating low
tone pattern, 10 subjects correctly notated G4 alternating from ear
to ear; 1 notated G3 alternating from ear to ear; and the subject
who had notated G5 alternating with G#5 for the alternating high
tone pattern notated G3 alternating with G#3 for the alternating
low tone pattern. All subjects except 1 therefore notated these
patterns correctly as consisting of the same pitch in both ears, and
the one exception notated a semitone difference between the tones
at the two ears.8The notations of an octave difference in listening
to the octave illusion pattern could not, therefore, have been due to
the subjects mistakenly attributing an octave difference between
the tones at the two ears.
This experiment provided more explicit documentation of per-
cepts of the octave illusion than have so far been obtained. It
should be noted that, because only 12 subjects were tested, the
experiment did not provide a measure of the statistical distribution
of the various percepts of the octave illusion, such as had been
provided earlier in the large-scale study of Deutsch (1974). How-
ever, the subjects in the present experiment were selected at
random with the only constraints being that they should be able to
read and write in musical notation, to have normal hearing, and to
be naive concerning the octave illusion. Eleven of the 12 subjects,
including those who notated complex percepts, notated tones at G4
and G5, and so notated them as separated by an octave; and 7 of
these notated the standard octave percept. One subject notated a
single pitch alternating from ear to ear, so that her percept fell into
the single pitch category described by Deutsch (1974).
In sum, 7 of the 12 subjects in this experiment notated the octave
percept of the octave illusion, and the notations of 4 more subjects
also included an octave difference between the tones, with only 1
subject notating a single pitch alternating from ear to ear. The
results of this experiment were therefore at variance with the claim
made by Chambers et al. (2002) that the octave percept of the
octave illusion is rare and that the perceptions of most subjects
involve pitch differences so small as to be amenable to an expla-
nation in terms of diplacusis. Rather, they supported the finding by
Deutsch (1974) that the majority of subjects perceived an octave
difference between the alternating tones—a finding that was also
in accordance with the report of Zwicker (1984).
The notated perceptions of the single tones of the same pitch
alternating from ear to ear provided a control for the possibility
8The pitch difference perceived by this subject when the same tone was
presented to the left and right ears may have reflected diplacusis. The
direction of this pitch difference did not correspond to the subject’s
patterns of localization for the higher and lower tones in the octave illusion.
Subject Characteristics and Notated Categories of Percept
Subject Age (yrs.)Handedness Percept of octave illusion Percept of alternating high tonePercept of alternating low tone
Single pitch (G4)
High tone (G5) in both ears
High tone (G5) in both ears
High tone (G5) in both ears
High tone (G5) in both ears
High tone (G5) in both ears
High tone (G5) in both ears
High tone (G5) in both ears
High tone (G5) in both ears
High tone (G5) in both ears
High tone (G5) in both ears
G5 in left, G#5 in right
High tone (G5) in both ears
Low tone (G4) in both ears
Low tone (G4) in both ears
Low tone (G4) in both ears
Low tone (G4) in both ears
Low tone (G4) in both ears
Low tone (G4) in both ears
Low tone (G4) in both ears
Low tone (G4) in both ears
Low tone (G4) in both ears
Low tone (G3) in both ears
G3 in left, G#3 in right
Low tone (G4) in both ears
“Handedness” refer to subjects’ familial handedness background: Right ? only right-handed parents and siblings; Mixed ? at least one mixed-handed
parent or sibling; Left ? at least one left-handed parent or sibling. For subjects who notated octave percepts, RR indicates notation of the high tone (G5)
on the right and the low tone (G4) on the left, with earphones placed both ways; LL indicates notation of the high tone on the left and the low tone on
the right, with earphones placed both ways; and Both indicates notation of the high tone on the right and the low tone on the left on one presentation and
the opposite localization pattern on the other. yrs. ? years; F ? female; M ? male.
All subjects listened to the octave illusion pattern twice, with earphones placed first one way and then the other. Notations in parentheses under
that the subjects, on listening to the octave illusion pattern, might
have mistakenly attributed an octave difference between the tones
at the two ears. The finding that the subjects notated the same pitch
at the two ears, with the exception of 1 subject who notated a
semitone difference between the alternating tones, is as expected
from other findings on diplacusis, in which the size of this effect
was generally found to be a small fraction of a semitone (see, e.g.,
van den Brink, 1975a, 1975b).
No correlate was here obtained between perception of the illu-
sion and the subjects’ handedness. This result was as expected
from the small number of subjects tested, given that large groups
of subjects are generally required for handedness correlates to
emerge at the perceptual level (Herron, 1980). Zwicker (1984) also
obtained no handedness correlates on comparing the perceptions of
3 right-handers with 3 congenital left-handers.
In this article, it has been argued that the study of Chambers et
al. (2002) used problematic procedures, so that their conclusions
concerning the lateralization of tones in the octave illusion were
called into question, as were their conclusions concerning the most
frequently perceived pitch differences between the alternating
tones. In addition, an experiment was reported that used a new
procedure that provided more explicit documentation concerning
the phenomenology of the illusion than in previous studies. This
experiment confirmed the observations in the original study of
Deutsch (1974) on the octave illusion, and its findings were
consistent with the two-channel model of the octave percept,
which invokes a separation of the what and where pathways in the
auditory system. The finding that the subjects notated an octave
difference between the alternating tones cannot be explained on
the proposal by Chambers et al. (2002) that the illusion results
from diplacusis. In addition, the diplacusis hypothesis cannot
explain the dependence of the illusion on tone duration or on
sequential context; neither can it explain the handedness correlates
with the type of percept obtained, which indicate that the illusion
serves as a reflection of brain organization, rather than character-
istics of the auditory periphery.
The two-channel model of the octave illusion has as its core the
supposition that the decision mechanisms underlying pitch and
lateralization are, at some point in the auditory system, distinct and
separate. This supposition is in accordance with recent findings by
auditory neurophysiologists. In particular, Rauschecker, Tian, and
colleagues have obtained evidence that the lateral belt area of the
auditory cortex of the rhesus monkey is subdivided into regions
that are specialized for the processing of either what or where
information: Neurons in the anterior belt are tuned specifically to
type of monkey call, whereas neurons in the caudal belt are instead
tuned to the spatial location of the signal (Rauschecker & Tian,
2000; Tian, Reser, Durham, Kustove, & Rauschecker, 2001).
Furthermore, it appears that information from these two regions
forms separate streams that project to spatial and nonspatial areas
of the frontal lobe (Romanski et al., 1999).
The two-channel model is in accordance with other perceptual
research indicating that different attributes of sound are processed
along separate pathways that are at some stage independent and so
can arrive at inconsistent conclusions (Carlyon, Demany, & Deeks,
2001; Darwin & Carlyon, 1995; Gardner, Gaskill, & Darwin,
1989; Hukin & Darwin, 1995; B. C. J. Moore, Glasberg, & Peters,
1986). Furthermore, Odenthal (1963); Efron and Yund (1974);
Hall, Pastore, Acker, and Huang (2000); Thompson (1994); and
Deutsch (1975b) have shown that illusory conjunctions of different
attribute values can occur with other sound configurations also.
The question then arises as to why people should have evolved
the two decision mechanisms that are hypothesized to produce the
octave illusion. This question can be addressed for the ear domi-
nance and lateralization components separately. Note that the ear
dominance component has two characteristics: First, it becomes
weaker as the duration between onsets of successive tones is
increased; second, it occurs in configurations in which the two ears
receive the same frequencies in succession and is weaker or absent
when this condition does not hold. Given these characteristics, it
was conjectured (Deutsch, 1981) that this effect reflects the oper-
ation of a mechanism that normally helps to counteract misleading
effects of echoes and reverberation. In normal listening, when the
same frequency emanates successively from two different regions
of space, the second occurrence may be an echo. This interpreta-
tion becomes less probable as the delay between these two occur-
rences is lengthened, and it becomes less probable when other
frequencies intervene between such two occurrences. On this line
of reasoning, the octave illusion falls into the class of phenomena,
of which the precedence effect is another example (Haas, 1951;
Wallach, Newman, Rosenzweig, 1949), which reflect the activity
of mechanisms that have evolved to counteract unwanted effects
due to the acoustics of the environment.
Concerning lateralization to the higher frequency signal, it was
conjectured (Deutsch, 1981) that this might reflect the action of a
mechanism designed to handle head shadow effects. When a
complex tone is presented in natural situations, the relative ampli-
tudes of the partials arriving at the two ears may differ consider-
ably, owing to the filtering action of the head. For example, when
a complex tone is presented to the listener’s right, then the higher
frequency components at the left ear are attenuated relative to the
lower frequency components. Assuming that the auditory system
interprets the pattern that produces the octave illusion as the first
and second harmonic of a complex tone, then it would make sense
to interpret the signal as coming from the ear receiving the higher
frequency—in this case, from the listener’s right.9
Finally, it should be emphasized that, in order to evaluate the
octave illusion, there is no substitute for listening to it as it was
originally generated. Furthermore, given the large differences be-
9An anonymous reviewer raised the question of what would be expected
on the two-channel model when the suppression component is weak or
absent. In the case of short tones separated by pauses, it is expected that the
normal process of harmonic fusion would take over and that the listener
would perceive a pitch that corresponds to the fundamental. This could also
explain the single pitch percepts obtained by some listeners. However,
when the tones themselves are of long duration, the process of fusion also
breaks down, and both the high and the low tones may be perceived and
lateralized correctly. A partial breakdown of both suppression and fusion
might also be responsible for the complex percepts obtained by some
listeners. In addition, depending on sequential context, both tones may be
perceived, but they may be incorrectly localized, as in the scale illusion
reported by Deutsch (1975b).
tween listeners in the way the illusion is perceived, it is important Download full-text
to have others listen to it also. The illusion occurs as a sound
demonstration in a number of publications, including Deutsch
(1983a, 1995); Houtsma, Rossing, and Wagenaars (1987); Kubovy
and Pomerantz (1981); and Pierce (1983).
Carlyon, R. P., Demany, L., & Deeks, J. (2001). Temporal pitch perception
and the binaural system. Journal of the Acoustical Society of America,
Chambers, C. D., Mattingley, J. B., & Moss, S. A. (2002). The octave
illusion revisited: Suppression or fusion between ears? Journal of Ex-
perimental Psychology: Human Perception and Performance, 28, 1288–
Darwin, C. J., & Carlyon, R. P. (1995). Auditory grouping. In B. C. J.
Moore (Ed.), Hearing (pp. 387–424). San Diego, CA: Academic Press.
Deutsch, D. (1974, September 27). An auditory illusion. Nature, 251,
Deutsch, D. (1975a, October). Musical illusions. Scientific American, 233,
Deutsch, D. (1975b). Two-channel listening to musical scales. Journal of
the Acoustical Society of America, 57, 1156–1160.
Deutsch, D. (1978). Lateralization by frequency for repeating sequences of
dichotic 400 Hz and 800 Hz tones. Journal of the Acoustical Society of
America, 63, 184–186.
Deutsch, D. (1980). Ear dominance and sequential interactions. Journal of
the Acoustical Society of America, 67, 220–228.
Deutsch, D. (1981). The octave illusion and auditory perceptual integra-
tion. In J. V. Tobias & E. D. Schubert (Eds.), Hearing research and
theory (Vol. 1, pp. 99–142). New York: Academic Press.
Deutsch, D. (1983a). Auditory illusions, handedness, and the spatial envi-
ronment. Journal of the Audio Engineering Society, 31, 607–618.
Deutsch, D. (1983b). The octave illusion in relation to handedness and
familial handedness background. Neuropsychologia, 21, 289–293.
Deutsch, D. (1988). Lateralization and sequential relationships in the
octave illusion. Journal of the Acoustical Society of America, 83, 365–
Deutsch, D. (1995). Musical illusions and paradoxes [CD]. La Jolla, CA:
Deutsch, D., & Roll, P. L. (1976). Separate “what” and “where” decision
mechanisms in processing a dichotic tonal sequence. Journal of Exper-
imental Psychology: Human Perception and Performance, 2, 23–29.
Efron, R., & Yund, E. W. (1974). Dichotic competition of simultaneous
tone bursts of different frequency: I. Dissociation of pitch from lateral-
ization and loudness. Neuropsychologia, 12, 249–256.
Gardner, R. B., Gaskill, S. A., & Darwin, C. J. (1989). Perceptual grouping
of formants with static and dynamic differences in fundamental fre-
quency. Journal of the Acoustical Society of America, 85, 1329–1337.
Haas, H. (1951). Uber den Einfluss eines Einfachechos auf die Horsamkeit
von Sprache [Concerning the influence of simple echoes on the intelli-
gibility of speech]. Acustica, 1, 49–52.
Hall, M. D., Pastore, R. E., Acker, B. E., & Huang, W. (2000). Evidence
for auditory feature integration with spatially distributed items. Percep-
tion & Psychophysics, 62, 1243–1257.
Herron, J. (Ed.). (1980). Neuropsychology of lefthandedness. New York:
Houtsma, A. J. M., & Goldstein, J. M. (1972). The central origin of the
pitch of complex tones: Evidence from musical interval recognition.
Journal of the Acoustical Society of America, 51, 520–529.
Houtsma, A. J. M., Rossing, T. D., & Wagenaars, W. M. (Eds.). (1987).
Auditory demonstrations. New York: Acoustical Society of America;
Eindhoven, the Netherlands: Institute for Perception Research.
Hukin, R. W., & Darwin, C. J. (1995). Comparison of the effect of onset
asynchrony on auditory grouping in pitch matching and vowel identifi-
cation. Perception & Psychophysics, 57, 191–196.
Kubovy, M., & Pomerantz, J. R. (Eds.). (1981). Perceptual organization.
Hillsdale, NJ: Erlbaum.
Moore, B. C. J., Glasberg, B. R., & Peters, R. W. (1986). Thresholds for
hearing mistuned partials as separate tones in harmonic complexes.
Journal of the Acoustical Society of America, 80, 479–483.
Moore, F. R. M. (1982). The computer audio research laboratory at UCSD.
Computer Music Journal, 6, 18–19.
Odenthal, D. W. (1963). Perception and neural representation of simulta-
neous dichotic pure tone stimuli. Acta Physiological Pharmacologica
Neerlanda, 12, 453–456.
Pierce, J. R. (1983). The science of musical sound. New York: Freeman.
Rauschecker, J. P., & Tian, B. (2000). Mechanisms and streams for
processing of “what” and “where” in the auditory cortex. Proceedings of
the National Academy of Sciences, 97, 1180–1186.
Romanski, L. M., Tian, B., Fritz, J., Mishkin, M., Goldman-Rakic, P. S., &
Rauschecker, J. P. (1999). Nature Neuroscience, 2, 1131–1136.
Thompson, R. F. (1994). Sensitivity to combinations of musical parame-
ters: Pitch with duration, and pitch pattern with durational pattern.
Perception & Psychophysics, 56, 363–374.
Tian, B., Reser, D., Durham, A., Kustove, A., & Rauschecker, J. P. (2001,
April 13). Functional specialization in rhesus monkey auditory cortex.
Science, 292, 290–293.
van den Brink, G. (1975a). Monaural frequency–pitch relations as the
origin of binaural diplacusis for pure tones and residue sounds. Acustica,
van den Brink, G. (1975b). The relation between binaural diplacusis for
pure tones and for complex sounds under normal conditions and with
induced monaural pitch shift. Acustica, 32, 159–165.
Varney, N. R., & Benton, A. L. (1975). Tactile perception of direction in
relation to handedness and familial handedness background. Neuropsy-
chologia, 13, 449–454.
Wallach, H., Newman, E. B., & Rosenzweig, M. R. (1949). The prece-
dence effect in sound localization. American Journal of Psychology, 62,
Zwicker, T. (1984). Experimente zur dichotischen Oktav-Tauschung [Ex-
periments on the dichotic octave illusion]. Acustica, 55, 128–136.
Received April 26, 2002
Revision received September 15, 2003
Accepted October 6, 2003 ?