In 2019, a student approached one of the authors of this article with a proposal for his senior project. He was a long-range shooting enthusiast and had recently purchased a suppressor for his rifle. Having used dosimeters and sound-level meters in class, he thought that characterizing the muzzle report of various firearms, with and without suppressors, would be a terrific way to put his newly acquired skills to work. Little did he know the surprise he was in for.

He selected seven firearms of widely different power cartridges, equipped them with some sound-level meters, and headed to the range. For each firearm, he placed the microphone one meter perpendicular to the muzzle. The results showed that the seven firearms were all nominally 141 dBA. If he placed the microphone seventy-five meters downrange to record the sonic crack, all were 143.3 dBA consistently.

Anyone experienced with firearms knows that the report of the diminutive .22 caliber long rifle is far quieter than the report of the .308 Winchester, yet the dosimeters gave the same reading. With the help of an insatiable curiosity, a pandemic lockdown, and a budding university research program, our quest to understand impulse noise was afoot.

HOW DOSIMETERS MEASURE SOUND Prior to using an instrument, industrial hygienists should have a good understanding of its mechanism of action—that is, how it works. This understanding should include upper and lower limits of detection or operation, the physical mechanism by which it operates, internal computations, and so on.

At 29 CFR 1910.95(d)(2)(i), OSHA’s occupational noise exposure standard states, “All continuous, intermittent and impulsive sound levels from 80 decibels to 130 decibels shall be integrated into the noise measurements.” Additionally, the standard’s appendix I requires “SLOW time response, in accordance with ANSI S1.4-1971 (R1976).” ANSI S1.4, Specifications for Sound Level Meters, states that the slow response is an exponential time averaging over one second. In 1971, when ANSI S1.4 was first adopted, memory for electronics was quite expensive; averaging the sound pressure once per second ensured that only 28,800 values needed to be stored to measure an eight-hour shift. Limiting the integration to values of 80 dB or greater reduced the need to carry digits of precision in the calculation, and limiting the upper range to 130 dB made microphones with good linearity and “toughness” relatively affordable. But this range was lower than the 140 dB impulse limit in 1910.95(b)(2).

Many manufacturers have since expanded the linear response range of dosimeters a little beyond the 140 dB limit. For sound pressures above the range of linear response, manufacturers have two options: implement a software cutoff that indicates the peak linear response value, or allow the instrument to report values outside of the linear range until peak voltage is reached and the system “clips”—that is, becomes overloaded. Given these options, how do dosimeters behave in an environment with a mix of impulse and continuous noise?

If a worker is in an environment with a steady 85 dBA background noise, and a 170 dBZ sonic crack occurs once every second and lasts 7 microseconds (µs), what results will the dosimeter provide in terms of peak and time-weighted average (TWA) values? Based on field testing, the peak reading would likely be around 157 dBZ and the theoretical TWA would be 85.2 dBA, assuming no reflections or reverberations. In a more realistic scenario of a 5 millisecond (ms) muzzle blast of 157 dBA at the ear of a shooter who is conducting endurance testing, firing once per second for an eight-hour shift in the same 85 dBA background would have a TWA of 111 dBA.

So, is your dosimeter lying to you about impulse noise? No, it’s doing what it was designed to do: measuring continuous noise to a reasonably good degree. However, your dosimeter isn’t telling the whole story when it comes to auditory risk. To understand the full picture, it is necessary to review the measurement of pure impulse noise and examine strategies to measure mixed impulse-Gaussian environments.

IMPULSE NOISE MEASUREMENT OSHA’s system for measuring Gaussian environments hasn’t progressed significantly since 1981, when the agency amended its occupational noise standard with requirements for hearing conservation programs. In contrast, the U.S. military has pursued better techniques for the evaluation of impulse noise through updates to MIL-STD-1474, Department of Defense Design Criteria Standard: Noise Limits. Developed in the 1960s and adopted in 1975, the “A” version of the standard set impulse noise limits. The “D” version adopted in 1997 added specifications for the necessary measurement equipment and analysis. The current “E” version improved on the measurement standards and a

dopted new hearing risk criteria for evaluating a single impulse. These criteria are the Auditory Hazard Assessment Algorithm for Humans (AHAAH) model and the LIAeq100ms metric.

As described in MIL-STD-1474E, the AHAAH model is “an electroacoustic analog of the ear structured to match the physiology of the ear, element for element.” The model calculates basilar membrane displacement and converts it to auditory risk units (ARUs), which are based on compound threshold shifts. CTS has components of both permanent and temporary threshold shifts. The standard limits exposure to 200 ARUs per 24-hour sliding window, not to exceed two such exposures per week.

The LIAeq100ms metric is the A-weighted equivalent sound level averaged over 100 milliseconds using a dosimeter’s impulse detector function. Similar in concept to the equivalent continuous sound level (Leq or LAeq8) that most industrial hygienists are familiar with, the LIAeq100ms metric is also readily converted to LIAeq8hr and from there to dose, so that each single impulse measured with and without hearing protection devices (HPDs) can be computed as the allowable number of exposures (ANEs) per day.

MIL-STD-1474E allows for use of either the AHAAH model or the LIAeq100ms metric. Both have adjustments for HPDs. MIL-STD-1474E specifies the use of ANSI/ASA S12.42 for the testing and rating of HPDs for impulsive peak insertion loss (IPIL), which is a measurement of how well HPDs attenuate impulse noise. However, most HPDs are not tested for impulse noise. While both risk models have their proponents and detractors for various technical reasons, a paper published in the International Journal of Audiology in 2019 found that for a broad range of firearm impulses, the computed risk scores between the two methods show a strong linear correlation, with an R2 of 0.99.

The most crucial element that MIL-STD-1474E provides is minimum specifications for measurement systems used for impulse noise. They must have at least 16-bit resolution, though 24-bit is preferred; a minimum sampling rate of 192,000 samples per second; and a microphone with a diameter not to exceed 6.4 mm (0.25 inch). These laboratory-grade systems can cost from $30,000 to $70,000 and are not particularly portable. For comparison, modern dosimeters typically have 24-bit resolution but sample at only 24 to 48 kilosamples per second (ksps); some handheld sound-level meters have sampling rates up to 96 ksps. High sampling rates are necessary to capture true peak impulse noise, which lasts approximately 7 µs. Small-diameter microphones are necessary since the pulse width of these impulses is only around 2.4 mm and will cause distortion across the membrane on a typical half-inch microphone, greatly distorting or diminishing the recorded peak value (PDF). While the physics and engineering of accurate measurement of strong impulse noises has been solved for the laboratory setting, the practicing industrial hygienist in the occupational environment is left bringing the proverbial knife to a gunfight.

MEASUREMENTS IN MIXED IMPULSE-GAUSSIAN ENVIRONMENTS Many industrial and manufacturing environments have a background noise level above the OSHA hearing conservation limit. This background noise from motors, conveyors, fans, and the like is described as Gaussian: normally distributed, with most noise centered around a particular level, say 86 dBA, and the distribution falling off in both directions in a classic bell curve. However, some industrial and manufacturing environments also have a strong impulse component that will, in theory, appear far to the right on the histogram—if the dosimeter is capable of accurately capturing the impulses. The histogram in Figure 1 demonstrates this bimodal distribution using a typical Type II dosimeter and OSHA hearing conservation (slow) settings. The large Gaussian peak centered around 76 dBA is from the fan or air noise at an indoor firing range, and the small peak centered around 127 dBA is from firearm noise. From the information presented in MIL-STD-1474E, we know the dosimeter hardware underestimates the impulse dose. What, then, is the industrial hygienist to do? Two strategies are emerging for the characterization of mixed Gaussian-impulse environments: kurtosis and an adaptation of the LIAeq100ms metric.