Recordings

Insertion Gain/Loss Measurement

To measure the earplugs we computed the difference at the manikin's eardrum between the open ear and earplug-inserted conditions (i.e., insertion gain/loss).

Stimulus: Uncorrelated pink noise presented from each speaker in the array such that a calibrated measurement mic (MicW i437L) in the ring center read 90 dB LAeq.

Insertion Protocol

To reduce any contribution of a particular earplug insertion, five separate recordings were made for each pair of earplugs. This generated a total of 10 recordings (5 insertions × 2 ears). Between recordings:

• Earplugs were removed
• Alternated between ears
• Reinserted

Hearing protection devices with a specific ear designation were only removed and reinserted (not alternated).

Each earplug was inserted in a manner intended to reflect a good seal at a realistic depth. In some cases this meant changing the size of the ear coupling to ensure a tight seal. Small amounts of water-based lubricant are used when there is friction between the earplug and the manikin ear canal.

The final computed metrics reflect the median value across the 10 recordings.

Diffuse Field Equalization

All recordings were diffuse field equalized using the same standard procedure described in our hearing device evaluation protocol.

Listening Simulations

These recordings were also used to create earplug listening simulations on HearAdvisor.com. To create these simulations:

1. A FIR filter was fit to the measured attenuation (earplug-inserted minus open) of the device, computed at ⅓ octave spacing
2. The filter was applied using a forward/reverse technique that does not add any delay
3. The output was used to create audio files in two versions:
   • (a) Unprocessed: No additional processing (demonstrates full earplug effect)
   • (b) Loudness-matched: Broadband linear gain applied after filtering to match the loudness of the open ear condition (allows the listener to focus on sound quality irrespective of loudness)