Whenever one measures the attenuation of hearing protection, it is different for every frequency. This has to do with the rule of acoustics – generally lower frequency, longer wavelength sounds do not “see” an ear obstruction such as an earplug or earmuff as well as the higher frequency, shorter wavelengths. Subsequently, lower frequencies have relatively little attenuation (typically no more than 25 dB at 250 Hz) whereas higher frequency sounds can be reduced by 30-35 dB. Having a single number to represent this non-uniform attenuation is like saying that a person with a mild to moderate sloping audiogram has a 20 dB (or should it be 60 dB?) hearing loss.
Single numbers are convenient, but sometimes we lose sight of some of the simplistic assumptions that are behind such a number. I frequently receive phone calls and emails from colleagues who want to know if a hearing protector with single number rating of 26 dB is adequate for a certain noise level (typically measured in dBA). My pat answer is I have no idea – if you send me a spectrum of the noise and the frequency-by-frequency attenuation of the hearing protector, I can do the subtraction for you. I am actually quite good at subtraction and received an A- in that subject in grad school.
Ideally, that is what we should be doing: subtract the frequency-by-frequency attenuations from the spectral levels at those corresponding frequencies. This will provide an estimate of what actually reaches the worker or musician and can inform us about whether the attenuation is sufficient, too much, or too little, and also at which frequencies corrections should be made.
For some reason, this was considered too difficult for some, so single number rating schemes were invented. There are several “single number” rating schemes in use around the world. In many parts of Europe the Single Number Rating (SNR) scheme is used. In North American the NRR takes the cake. The NRR is the Noise Reduction Rating and is now about 35 years old.
This is how it works…
The NRR is the noise (C-weighted) – noise (A weighted) + attenuation (A-weighted) – 3 dB – 2 standard deviations.
Don’t panic! If there is minimal low frequency spectral energy below 1000 Hz such as in many forms of treble music, the first two terms cancel out (noise [C-weighted] – noise [A-weighted]), so we are left with the actual attenuation and then we get to subtract off estimates of how poor the hearing protection can actually be if worn poorly. This last series of correction factors is actually quite brilliant and seeks to estimate a worst-case scenario.
If the hearing protector was uniform in its attenuation, such as the ER-15 Musicians Earplugs, the NRR formula would yield an unfairly low number for the earplug (because the attenuation would be measured in A-weighting). The ER-15 attenuates all sound by 15 dB but it has an NRR rating of about 10 dB.
To be accurate, however, the Environmental Protection Agency (EPA) in the United States requires the NRR number only for non-custom products. So, let’s take another musicians’ earplug that has a relatively flat or uniform attenuation – the ETY from www.etymotic.com. This one-size-fits-all earplug (actually, it’s now two-sizes-fit-all: a standard and a large) has an attenuation characteristic that ranges from about 18 dB for low frequency sounds to about 22 dB for high frequency sounds. The associated NRR would be about 12 dB. There is nothing “12 dB” about the ETY earplugs.
Also, if the musician was instructed properly on how to use it, the poorest fitting would not be as poor as a one-size-fits-all earplug in an industrial environment., and the penalty of “3 dB – 2 standard deviations” would not be as great.
Next week, part 2 will be a whimsical overview of how we may use a 2 number rating scheme that would be more useful for describing the attenuation characteristics of custom and non-custom hearing protectors.