If you have been reading earlier parts of this blog series, I had only promised to write a 2-parter.  Well, here is part 3!

The reason for this rogue insertion of yet another installment to this series is the number of emails that I have received about how hearing protection can actually be assessed using real ear measurement – hence, part 3.

I have written about this in earlier blogs but it’ best to have all of this in one place. Using real ear measurement to assess the extent and frequency dependence of hearing protection is quite similar to using it for assessing gain and output with hearing aid amplification. 

Courtesy of www.Audioscan.com

There are however two (potential) differences. 

One is that most modern hearing aids are “level dependent” where different amounts of gain are generated depending on the level of the input – hearing aids generate more gain for soft sounds and less for louder sounds. There are level dependent hearing protection devices and these should be assessed using the identical protocols for level dependent hearing aids, but these are rather rare in the realm of industrial or music exposure. The vast majority of hearing protection devices are linear and not level dependent. 

The other difference is that with hearing amplification, the output is at a relatively high level; certainly higher than the internal noise floor of the real ear measurement device. This is not the case with many forms of hearing protection devices such as Musicians’ Earplugs. 

For example, if a 50 dB SPL stimulus level was to be used and the hearing protection device provided 30 dB of attenuation, then the output would be on the order of 20 dB SPL. The vast majority of real ear measurement systems have a noise floor that is greater than 20 dB SPL.

One can easily measure the noise floor (or low sound level limit) of their own real ear measurement system by doing the following:  Calibrate the real ear measurement system in the normal fashion and then disable the loudspeaker and reference microphone.  Different manufacturers have differing terminologies for this – with Audioscan setting  the stimulus level to 0 dB, and Frye setting the stimulus level to “off”.  In both cases, the loudspeaker and the reference microphone are both disabled.

Courtesy of www.iso.org

Plug up the end of the probe tube either by pinching it between your fingers or using a pair of plyers.  Perform a “run” of the real ear measurement system as you would for a hearing aid. The result is a frequency by frequency measurement of the internal noise floor of your own real ear measurement system. One needs to be at least 10 dB above the noise floor in order to be able to trust your results.

All this means is that instead of using a 50 dB SPL stimulus level, one needs to use a higher level one. I use 70 dB SPL as my stimulus level for assessing the real ear attenuation of hearing protection devices. Of course one needs to be circumspect about the low frequency results due to potential slit leaks between the hearing protection device and the ear canal wall caused by the presence of the probe tube.  Using Vaseline or other “gooey” substance around the outer edges of the probe tube will resolve this potential issue.

Using a 70 dB SPL stimulus level will ensure that the output measured by the real ear measurement system will be far above the noise floor.

In part 1 of this blog series, I had talked about a patient who had a relatively rare 3000 Hz noise (music) induced hearing loss notch and not the more common one at 4000 Hz or 6000 Hz.  The ear canal resonance was at 2000 Hz as compared as opposed to the more common “average” of around 2700 Hz.  The lower frequency outer ear resonance was indeed a major reason why there was a 3000 Hz notch rather than in the 4000-6000 Hz region.  

There are several theories of why we have a noise induced hearing loss in the 3000-6000 Hz region but one of them is related to having the greatest hearing loss about one half octave higher than the offending frequency.  If music (or noise) is enhanced in the 2000 Hz region prior to even getting to the cochlea, then one half octave above this would be 3000 Hz, so this finding is not that surprising (like a G being a half octave higher than the C below it).

What is surprising is finding a 2000 Hz outer ear resonance- this guy must have an ear canal that is 10 feet long!

The question introduced in part 1 of this blog series is what do we do when we want to ensure a flat or uniform attenuation with hearing protection such as Musicians Earplugs.  These earplugs “assume” a 2700 Hz resonance in order to create a flat response.   Do we just “take the hit” and put up with a 2000 Hz greater attenuation (as a result of the destruction of this person’s natural ear canal resonance) or do we modify the Musicians Earplugs in order to obtain a 2000 Hz resonance which can be off-set by this patient’s low frequency outer ear resonance?

Smoothness or good high frequency acuity? Courtesy of www.devot-ee.com

The short answer is “I don’t know” but the slightly longer answer is, let’s find out.

Two pairs of custom silicon earmolds were made for this patient both with the ER-15 Musicians’ Earplug.  In the manufacture of these earplugs a “mass meter” is used to ensure that the volume of air in the earplug bore resonates with the known compliance of the ER element.  This is known as a Helmholtz resonance- an interplay between volume of air and compliance creates a resonance.  (There is another way to do this with a quarter wavelength resonator and this is the idea that is used in the non-custom one-size-fits-all ETY earplug or ER20XS earplug).

But back to the Helmholtz resonance used in the custom earplugs.

One can alter the resonant frequency of these earplugs by changing the dimensions of the bore of the custom made earplug.  It can be made narrower or wider, longer or shorter, than prescribed and verified by the mass meter readings.

The Musicians Earplug has a silicon custom earplug with an acoustic pathway whose dimensions can be altered. Courtesy of www.Etymotic.com

Another Musicians Earplug approach, with this particular model being a one-size-fits all. Courtesy of www.worldwidemusic.co.uk

Specifically, the resonance frequency is proportional to the cross-sectional area and inversely proportional to the length.  Increasing the bore width and decreasing the bore length will increase the resonance frequency, but doing the converse (a long, narrow bore) will create an earplug that has a lower resonant frequency (around 2200 Hz).

Subsequently two earplugs were made- one with normal dimensions and another with a longer narrower bore.  The longer and narrow bore one would provide a flat response in the in 2000 Hz region and would continue to be smooth up to around 3500 Hz with a roll-off (greater attenuation) above that.  The conventional bore dimensions would result in a notch in the 2000 Hz region, a slight bump (less attenuation) in the 3000 Hz region, but otherwise uniform attenuation above that up to about 8000 Hz (my office probe tube microphone doesn’t measure above 8000 Hz).

And the winner is….  That is, the preferred earplug arrangement for this one patient after trying it for several weeks, is…. The longer narrower bore that resulted in a smooth response up to about 3500 Hz even in the 2000 Hz region.  He preferred that over having better high frequency acuity but a notch in the 2000 Hz region.

I am not sure that a single subject using a single subject design is  definitive, but as stated in part 1 of this blog series, this would make for a really interesting and clinically practical AuD Capstone study.  The student could use a simple equalizer to alter the frequency response heard in a number of different ways (including a 2000 Hz notch) and play these audio files to a large number of subjects.  And then I would hire that student!