ï~~OTOACOUSTIC EMISSIONS AS A COMPOSITIONAL TOOL
Jonathon Kirk
Eastern Illinois University
Department of Music
ABSTRACT
This paper describes various compositional attempts at
initiating otoacoustic emissions through electronic music
composition. Otoacoustic emissions (OAE) are sounds
that arise in the inner ear and which are related to the
process of amplification in the cochlea. While the
physiological processes involved with OAEs are not
entirely understood, scientists now have a better
comprehension of how they work within the inner ear.
This paper will provide a basic overview of OAEs along
with their compositional applications by assessing
particular works of Maryanne Amacher, Jacob Kirkegaard,
and Phill Niblock.
1. INTRODUCTION
Otoacoustic emissions (OAEs) are physical sounds that
arise in the ear canal when the tympanum receives
vibrations transmitted backwards through the middle ear
from the cochlea [2]. While aural harmonics and
subjective distortion products have been understood as
psychoacoustic phenomena since the eighteenth century,
by the mid twentieth century there had not been an
adequate explanation to account for a physical appearance
of these tones in relationship to the processes of the inner
ear. OAEs were first predicted by NASA scientist Thomas
Gold in 1948 and later discovered by auditory physicist
David Kemp in 1977 [10]. Gold based his cochlear model
on his experience with radio technology and the
transduction of frequency information by regenerative
receivers. Kemp followed Gold's model by suggesting
that distortion products could be generated within the ear,
and that these products could be described as 'evoked
cochlear mechanical responses' or 'echoes' [7]. Scientists
then began to infer connections between perceivable
internal oscillations in the ear and physical emissions from
within the inner ear.
Eventually these responses would be described as
OAEs, and it would be shown that the sound produced by
the ear could be measured and recorded in the human ear
canal. With the knowledge that the inner ear could indeed
generate sound, Davis (1983) described the presence of a
'cochlear amplifier' and it was later demonstrated that the
outer hair cells were capable of receiving and transducing
energy [3]. Ultimately it was shown that OAEs could be
generated as distortion products by the application of two
sinusoidal stimuli, and it was also recognized that one
could record OAEs by placing a small microphone inside
the ear canal. As the reality of OAEs were confirmed and
the processes of the inner ear became better linked to
physical sound, we now understand that the cochlea uses
active processes to detect the phenomenally faint sounds
(measured in micropascals) that our ears routinely hear [2].
2. OAE CATEGORIZATION
OAEs are generally categorized in three types:
spontaneous otoacoustic emissions (SOAE), transient
evoked otoacoustic emissions (TEOAEs), and distortion
product otoacoustic emissions (DPOAEs). DPOAEs are
generated by the simultaneous presentation of two closely
spaced sinusoidal frequencies where the two frequencies
activate the cochlea in the same region of the basilar
membrane [5]. It was implied early on that OAEs were
consistent with the appearance of subjective combination
tones and in some cases with the resultant sounds caused
by diplacusis, but it was shown by Kemp (1977) that the
mechanical process of intermodulation distortion parallels
the stimulus parameters involved in the generation of
OAEs: the quadratic difference tone (QDT) appears at
frequency fQDT 12 - 1f where f2 > 1f. While other
distortion product (DP) frequencies are possible of being
generated, the strongest additional DP in humans is
defined by the cubic difference tone (CDT), fCDT= 2f, -
f2. While nonlinear distortion should generate the sum
tone as well, it is much more difficult to hear because it
can be masked by the original tones. Furthermore the
ordinary difference tone can be difficult to detect when it
lies between fI and f2 [4]. Emergence of DPOAEs in the
cochlea show that when there is nonlinearity anywhere in
the sound transmission path it should generate an
additional sounding element.
SOAEs, faint, pure sounds from within the ear, were the
first to be reported and are generally used in the testing of
infant hearing, while TEOAEs are similar to DPOAEs and
are recorded in response to a very abrupt (click) stimulus
[5]. Because acoustic stimuli are easily controlled and the
resultant tones are mathematically predictable, DPOAEs
are perhaps the most advantageous to record and measure.
316
