Audiometric notches at different frequencies

Have you ever noticed when testing a musician or even a teenager who walked into your office that the patient has a slight 4000 Hz notch in one ear and a 6000 Hz notch in the other ear.  From a strictly “asymmetrical hearing loss” perspective, this may fulfill the requirements for a retrocochlear investigation, which should be pursued. But I wonder if there are some acoustical reasons for this finding.

The use of earphones in music is one of the salient differences between hearing loss among musicians and hearing loss among their industrial colleagues.  Workers in a factory primarily wear earplugs to attenuate the sound.  Musicians wear earplugs as a conduit to hear amplified sounds (via an insert earphone).  Industrial workers are subject to long-term hearing loss from sound that still gets past (and around) their hearing protection, as well as from not wearing hearing protection properly.  Musicians and also us mere mortals who just like to listen to music are anything but accepting of earplugs that do not fit well.

A poorly fit earplug allows the lower frequency sounds to get into the ear (in an industrial environment) and to leak out of the ear (in a musical environment) and, despite our best efforts (as well as those of our industrial friends), earplugs are not ideal.

The human ear canal is a dynamic structure whose anterior (‘front’ for those musicians reading this) portion moves forward and backwards as we open and close our mouths.  CT and MRI studies have shown that as we open our mouths, there is no significant change in the rear (posterior) or upper and lower (superior and inferior) dimensions. It is only the front direction that changes.

When a teenager is walking down the street, chewing gum or sipping a latte (if they live in a large urban center), his jaw slides forward and backward.  Despite the fact that their earplugs holding their MP3 player earphones are soft and malleable, acoustic slit leaks can occur between the earplugs and the front of their ear canals.  When this happens an entire host of other things can happen as well.

With a slit leak, the low-frequency end of their music may be lost to them; they may need to turn up the volume on their MP3 to compensate for this lost “low end”; and the spectral shape of the music may be different.

To make it even more difficult to study, one ear may be receiving different levels and spectral shapes of music than the other ear.  Any dentist will tell us that our jaws (and temporalmandibular joints- TMJs) are not symmetrical.  If we have a TMJ dysfunction, it typically is worse on one side than the other.

This, as well as the specific shape of our ear canals, may be reasons why the sound level of the music that we are subject to may differ from one ear to our other ear.  It may be that due to the twists and turns of our ear canals we are able to insert an earbud or earplug only so far and the distance may differ between our ears.

When there is a different insertion depth and different amounts of loss of the lower frequency sound energy, the ultimate spectral level and spectral shape of music may differ from ear to ear.

Clinically I tell my patients who like to listen to MP3 players and other forms of portable music to open their mouths when they insert their earbuds.  This allows the jaw to slide forward, thereby allowing them to insert the earplug deeper into the ear canal.  (Incidentally this is a great clinical trick for audiologists who are having difficulty obtaining a proper hermetic seal during the admittance testing portion of a hearing test. Having patients open their mouth allows the probe tip to be inserted deeper with an improved chance of an adequate hermetic seal. It also stops them from talking during the tympanogram!… better, and more professional than having to say, “Please shut up!  Your talking is screwing with this test!”)

We are not actually sure why we have an audiometric notch at the frequency that we do.  There are many theories about why it is at 3000 Hz, or 4000 Hz, or 6000 Hz, and this may be related to the location of the peak in the ear canal; permanent hearing loss tends to show up about ½ an octave above the peak.  In the unoccluded ear canal, the natural resonance is around 2600-3000 Hz, so it stands to reason that 4000-5000 Hz (being about half an octave higher in frequency) is where the permanent damage occurs.

The situation becomes more complex in the occluded (or semi-occluded) ear canal.  Acoustically we are not sure what typically happens with MP3 type earbuds on an individual basis.  (HOWEVER THIS IS EMPIRICAL AND WOULD MAKE A FASCINATING CAPSTONE PROJECT FOR A STUDENT).  Understanding more about the acoustics of the occluded ear may bring us closer to understanding why there may be slight differences in damage between a person’s two ears and why the maximum audiometric loss may be at different audiometric test frequencies.

About Marshall Chasin

Marshall Chasin, AuD, is a clinical and research audiologist who has a special interest in the prevention of hearing loss for musicians, as well as the treatment of those who have hearing loss. I have other special interests such as clarinet and karate, but those may come out in the blog over time.

1 Comment

  1. One theory involves perilymphatic fluid flow and turbulence past the basal turn of the cochlea (located at approximately 5-6mm from the oval window), at the points approximately 8-10mm in. See the Greenwood chart at:

    It is further posited that the shear forces from this turbulence act to literally rip out the outer hair cells in this region.

    There are basically three states of this fluid flow: Laminar flow when the Reynolds Number Re is under about 2400, turbulent fluid flow when Re is greater than 4000; and the transition region when Re is between 2400 & 4000. The Reynolds Number is a dimensionless number that gives a measure of the ratio of inertial (kinematic) forces to viscous forces for given gaseous and liquid fluid flow conditions.

    To determine the actual fluid flows inside the cochlea, one would first model the inside of the organ and then solve the Navier-Stokes Equations:
    …using finite element analysis software, such as COMSOL Multiphysics.

    For more on the Reynolds number, please see for a visual demonstration:
    and for the equations, please see:

    For more on the Navier-Stokes Equations, please see:

Comments are closed.