How We Test Earplugs

Evaluation of earplugs is hard and no methodology is perfect. We’ve carefully considered many audiological and acoustical decisions. Our process and rationales are described below and, given our methods, our data better represent a “best fit” model as a manikin was used for measurements and earplugs were fit by an audiologist.

How do you record the earplugs?

Hearing protection devices are fitted to a KEMAR acoustic mannequin⁸. The manikin uses anthropometric pinnas and VA-style tapered ear canals. The quality of the physical fit is assessed by measuring the occluded response and comparing plots to manufacturer data. Once fit is deemed relatively accurate, five occluded sine sweeps are recorded with the earplug completely removed and reinserted between recordings to account for fit variability. Music sound samples are then recorded once the occluded responses are symmetric between the manikins ears and matching expectations for occlusion.

KEMAR is located in a quiet acoustically-treated room, surrounded by a ring of 8 speakers (1m radius). Sine sweeps are presented over the 0 degree speaker using Room EQ Wizard (REW) and recorded from the eardrum reference position of the manikin. To make the recordings appropriate for presentation over commercial headphones, all recordings were post-processed via a diffuse field equalization filter⁵ customized to the test environment (see Figure 1).  Lastly, all recordings are filtered to remove frequencies above 10 kHz due to step half-wave resonances present in the GRAS RA0045 ear simulator.

How are you sure the testing room is appropriate?

The testing room was custom built to be appropriate for the assessment of ear level technologies. The room is sufficiently quiet and non-reverberant having an RT60 of 64.2 milliseconds. This reverberation time corresponds to a critical distance, or distance where the direct path energy equals that of the reverberant, of 1.37 meters. Therefore, the 1 meter radius in our laboratory is well within the critical distance, ensuring that our multichannel reproduction is dominated by the direct path from speaker to manikin .

How do you present the music?

We simulate acoustic music listening by playing several genres of recorded (royalty-free) music from a stereo pair of speakers located at -45 and +45 degrees.

How do you calculate Sound Quality

Our Sound Quality metric is subjective and corresponds to how well the earplug preserves musicality when worn. If an earplug sounds as though the world is equally “turned down” it receives a higher score. Devices that perform more similar to industrial plugs, creating a muffled sound quality, will receive a lower score. This metric is similar to our Uniformity rating which objectively assesses the flatness of sound attenuation.

While this rating is subjective, there is a consistent process performed for each device. First, the earplugs are inserted and music is presented from a stereo speaker system at roughly 90 dBA. Sound quality is also assessed by strumming chords on an acoustic guitar and conversing in a one-on-one situation. In all cases, the overall balance between high and low frequencies is compared to normal listening without earplugs. A rating is assigned once testing is completed.

How do you calculate Sound Protection

Sound Protection is an objective rating based on the manufacturer listed Noise Reduction Rating (NRR). A few calculations are performed to estimate the duration of time someone may be protected while wearing each device at a common concert level of 100 dBA. These estimates assume no other hazardous exposures throughout the day and are derived using the following steps:

First the protected exposure level is estimated using a 7dB correction for dBA measures and 50% derating per OSHA:

• Protected exposure (dBA) = Unprotected exposure (dBA) - [0.5(NRR-7)]

Then the maximum allowable time at that given exposure is estimated using the recommended exposure limits (REL) outlined by the National Institute of Occupational Safety and Health (NIOSH)¹⁰:

• Time (hours) = 8 / 2 ^ (Protected exposure (dBA) - 85)/3

The final number corresponds to the amount of time, in hours, an individual can be exposed before auditory injury occurs per NIOSH recommendations. For reference, 15 minutes is the maximum duration for a 100 dBA exposure level.

The following table is used to compute a final metric for Sound Protection:

A one hour ceiling was used as the high-end reference for several reasons. Mainly, it is  a logical period of time that is easily conceivable for live events. Secondly, choosing a longer concert duration, or louder overall sound pressure level, portrayed most earplugs as insufficient protection. As we do not wish to convey that earplugs are ineffective, a one hour standard is kept.

The greatest available NRR was used to assign a rating to devices offering various filters or adjustable attenuation.

How do you calculate Our Reduction Rating

Our Reduction Rating is another objective metric associated with sound reduction. The process outlined above for the Sound Protection metric is performed again based on the insertion loss data collected in our lab. To obtain insertion loss, sine sweeps are recorded through the manikin with and without earplugs inserted. Five sweeps are performed with earplugs being completely inserted and removed between recordings to account for insertion variability. The data from each condition are averaged and subtracted from one another to obtain the degree of attenuation provided, i.e. insertion loss:

• Insertion Loss = Open Ear Response(AVG) - Occluded Ear Response(AVG)

NRR is calculated as specified in ANSI S3.19-1974¹ and the remaining steps mentioned above for the Sound Protection metric are performed including the table for providing a final rating.

We have opted to refer to this as “Our Reduction Rating”, as opposed to a secondary NRR, as our anthropomorphic manikin may not accurately replicate the many complexities found in human anatomy and hearing⁷. So while this metric may not accurately reflect real-ear attenuation, intra-device comparisons regarding insertion loss can provide useful information as the test setup remains constant.

How do you calculate Uniformity

Our Uniformity metric corresponds objectively to the flatness, or uniformity, of sound attenuation provided by each earplug. We deem this metric of high importance as it directly relates to the underlying assumption of musicians and high-fidelity hearing protection devices. That is, a flatter profile with equal sound attenuation should translate to a more natural and musical sound quality.

Flatness is calculated by expressing the difference in average dB attenuation between 250 and 4000 Hz over the corresponding octave ratio^{11, 12}:

• Slope = | 250Hz(AVG)- 4000Hz(AVG) | / 4 Octaves

The slope of all devices is then organized into five equal tiers from least (flattest) to greatest and a score from 5 to 1 is assigned, respectively. Slope scores are averaged to obtain a single overall rating for devices that offer multiple filters or adjustable attenuation.

How do you calculate Overall Value

Overall Value is a subjective metric based on the collective perceived quality of a device and its offering. All elements of the fit are weighted against both professional audio and clinical experience to reach a final rating.

How do you calculate Fit and Comfort

Fit and Comfort is a subjective rating based on user experiences during testing. It encompasses ease of insertion, earplug retention in ear, short term comfort while conducting the Sound Quality metric, long term wear comfort (at least 1 hour of continual use), and comfort during extreme jaw movements such as chewing and singing.

The best style and size ear tip available is used for testing and rating. For devices that ship with only one size ear tip, the option believed to be most appropriate is selected at the time of purchase.

How do you calculate Occlusion

Occlusion is a subjective rating based on user experiences during testing. Occlusion refers to a perceived “hollow” or “boomy” voice quality when the ear canals are plugged. Higher scores correspond to devices with minimal occlusion effect as it is generally regarded as undesirable. Devices exhibiting an increased sense of isolation and voice resonance receive a lower rating. Occlusion is assessed by speaking, singing, and chewing while earplugs are worn.

How do you calculate Price

Price can be an influencing factor when selecting a hearing protection device. While most earplugs are relatively inexpensive, we have seen higher costs associated with some of the newer adjustable options. A simple 5 tier rating system is used to accommodate this range:

How do you calculate Build Quality

Our Build Quality rating is subjective based on handling and use with each earplug. The perceived quality of the earplug machining, build materials, filters, and silicone ear tips are considered. It is also possible for two earplugs of the same model to exhibit slight performance differences. This may suggest lower quality control and less consistency between acoustic filters in the manufacturing process. This is weighted in this rating if deviations are observed while testing. 

How do you calculate Extra Features

Extra Features is a direct rating of any additional supplies offered with an earplug. While seemingly secondary to elements like sound quality, accessories such as multiple ear tips and a keychain carrying case can have a significant impact on earplug use. We compiled some of the more common and desirable features into the following table and tally the Extra Features score:

How do you calculate Visibility

Visibility is a subjective rating of how noticeable an earplug is when being worn. Devices with larger profiles that protrude from the ear canal receive a lower score while smaller, transparent, deeper fitting devices are rated more favorably. This metric is potentially controversial as individual preferences for earplug style will vary. However, we see a general consensus towards more inconspicuous ear technology and based this metrics scale on this trend. 

Disclaimers

Measurements were obtained on KEMAR which is one of the most well documented head and torso simulators for research and development⁸. No manikin to date however can perfectly replicate complexities of the human head and its associated flanking paths. Therefore, the limit that we considered when determining the validity of our measurements is the case of a perfectly blocked ear. We measured this by blocking the manikin ear canal with putty to obtain a self-insertion loss measurement⁹. In this condition, any sound arriving at the ear canal microphone is not through the traditional air conduction path—the path of interest for earplug performance. The attenuation in that condition is the lighter-gray curvy line visible in Figure 1. Any values below this line are not valid because we are not able to distinguish the air conducted (target) path from the bone conduction (interference) path. In humans, the limit is the bone conducted path (see the blue line with square markers in Figure 1)^{6, 7}. The entire gray region indicates where out measurements may not reflect the acoustic isolation of a human head.

Those in hearing sciences may reference the preferred approach for measuring earplug insertion loss which involves human subjects and a real-ear attenuation at threshold (REAT) method⁴. While this is the “gold standard”, it poses many practical limitations. An acoustic test fixture, as we have used, allows us to perform consistent repeatable measurements and every effort is made to provide viable information given the limitations discussed.