When you think back 50 years to the noise measuring instrumentation of the 1970s, they seem rather primitive by today’s standards.  Speaking of which, IEC ^[1] and ANSI standards ^[2] exist to ensure measurement accuracy and consistency across different manufacturers; and that’s really important when you think that the outcome of your quantitative risk assessment according to methods like ISO9612 ^[3] or OSHA 29CFR 1910.95 ^[4], means controlling employees noise exposure through a hearing conservation programme and improved hearing health outcomes.

Rather like mobile phones, which have reduced in size and increased in functionality, conceptually there is little to choose between a professional hand-held sound level meter (SLM) and a bodily worn noise dosimeter other than form factor, notwithstanding that they have different applications.  Both types of instruments have tremendous processing power offering the capability depending on the exact model to measure multiple parameters in parallel, assorted averaging and time profiles, real-time frequency analysis, audio recording, the list goes on. And even with all this functionality, they couldn’t be easier and more intuitive to use. But despite this progress there are some basics that have remained constant.  Let’s review them.

Prior to the digital revolution of the early 1980s which witnessed the application of microprocessors and CMOS memory and then again in the early 1990s with A-D converters and digital signal processing (DSP), the very original SLMs were literally not much more than a voltmeter, with a microphone and preamplifier attached.  The familiar ‘pointed shape’ is a result of the need to stop reflections from the instrument’s case that can otherwise cause measurement errors. Noise dosimeters by contrast were little more than a cigarette pack sized box with a microphone on a cable. Modern dosimeters now have a microphone close coupled to the body of the instrument which removes the vulnerable cable but is no mean feat to achieve in acoustic reflection terms. Indeed, the use of noise dosimeters remains the preferred method in the US, but they have been less prevalent in Europe due to a preference for an SLM which stems from the fact that the risk of noise induce hearing loss (NIHL) was based on SLM results going back to the 1950s. 

SLMs then used a commercial off the shelf RMS detector with hardware defined time weightings e.g., Fast (F) or Slow (S), maybe Impulse (I), to display sound pressure level (SPL) on a moving coil meter which was ‘damped’ by the time weighting and scaled in decibels (dBs).  The dynamic range over which the instrument could measure may have only been 30 or 40dB requiring several potential range changes but its common now to get 20-140dB in a single span.  Formerly you could easily have used incorrect settings but today there are pre-defined setups that can be chosen to satisfy local workplace noise legislation to minimise human errors.

Original dosimeters simply displayed percentage dose with one chosen Q-factor and criterion which determines how much noise constitutes a doubling of energy and an allowable 100% dose respectively.  Q=3 is the norm in Europe, the so-called equal energy principle, whereas 4, 5 & 6 are used in the US depending on Government Agency, although NIOSH now favour the more stringent European approach to a doubling of energy and hence risk.

Let’s not forget frequency weightings including the now familiar A & C or Linear (L) weighting.  Linear had never actually been standardised and varied between manufacturers, which is why you now often see a Z-weighting rather than Linear in current instruments.  These weighting curves were based on original research work by Fletcher & Munson ^[5] from empirical human studies way back in 1933 on the perception of loudness; those curves in red in Figure 1 are from the latest ISO standard ^[6].  The curves essentially flatten with increasing level, and you can see the similarity with the current weighting shapes in Figure 2 albeit you have to imagine them in reverse.  At one time you were required to switch between weightings A, B, C & D depending on the level but that proved rather impractical, and A & C became the norm with B & D long forgotten for use in the workplace. 

Figure 1

Figure 2

Frequency Analysis

Frequency analysis is extremely useful for the most accurate selection of hearing protection, although there is a simplified method based on a C-A weighted value calculation.  Modern DSP provides simultaneous A, C and octaves (or optional 1/3rd octave) band analysis whereas sequential frequency analysis was once achieved via a ‘strap on’ box making the whole thing quite heavy and time consuming having to manually step though 10 octave or 33, 1/3^rd Octaves!  And yes, there is a standard for the frequency filters ^[7].

Nothing wrong with Average.

Some instruments in the late 1970s might have had a direct reading of the equivalent average value (Leq), a major innovation at the time but now standard on all but the most basic of instruments.  Prior to direct reading Leq it was a case of taking an eyeball average of the changing display, set to a Slow time constant, a change of less than +/- 4dB could be taken as the Leq.  The use of a Slow time constant persists in the US to this day. 

All of these weightings and time constants are now synthesised in embedded software running in a DSP and regardless of the setup used, instruments measure and store all relevant parameters even if not selected for display which can also be viewed, if necessary, in associated software.

Accuracy

Historically, there were four grades of instruments namely 0, 1, 2 and 3; the former for laboratory-based measurements, then precision, industrial and indicator respectively.  Precision and industrial grades loosely compare to the Class 1 & 2 that exist today, their performance largely dictated by the frequency response and sensitivity of the microphone. Instrument designers apportion 2/3^rds of the allowable frequency response tolerance to the microphone and 1/3^rd to the electronics because the former can drift with temperature and humidity effects.  Therefore, noise measurements are unique in requiring calibration before and after use. And even that process is now automated with the calibration signal being detected and adjusted automatically; gone are the days of ‘twiddling’ a potentiometer with a small screwdriver. 

However, calibration is something of a misnomer because the act of putting an acoustic calibrator on an instrument is more correctly a field check, normally with one sound level, typically 114dB, which means you can calibrate in a noisy environment and with one frequency level, typically 1KHz. This only checks the instrument at one measurement point and ignores linearity, impulse and frequency response and other vital characteristics.  Its therefore advisable to send your instrument back for a manufacturers certificate of calibration every two years (or as local legislation dictates) and that includes the calibrator itself.

Type 1 microphones are pieces of precision manufacturing, many of which get rejected in manufacture due to very tight sensitivity, capacitance, and frequency response requirements, which means they are consequently quite expensive.  By contrast type 2 microphones have a narrower frequency response and wider tolerance compared with type 1 and are therefore simpler (and cheaper) to manufacture.

Thinking about the earlier smartphone reference of course ‘there’s a noise app for that’. There are many available and NIOSH in the US ^[8] developed an app with the aim of democratising noise measurements. However, its only IOS based, because Android phone performance is too variable which may impact on accuracy, and they further recommend that they only be used with an external microphone suitable for checking with an acoustic calibrator. A sound level meter, right?  And, of course, there are many places where phones are banned for security reasons or due to hazardous atmospheres.  Hearing protectors with built-in noise monitoring are also on the increase but there is something of a catch-22 in their use; to meet the recently revised PPE regulations ^[9] as a Category III device they ought to be type approved against the appropriate standard for noise dosimetry but conversely there isn’t a standard to cater for this set-up yet. Category III products are designed to offer protection against more serious hazards, which have the potential to cause serious harm or death, and reinforces the seriousness attributed to NIHL.

By contrast, dedicated noise dosimeters are readily available that meet the relevant instrument and application standards plus they have intrinsic safety approval for use in petrochemical, pharmaceutical and mining applications. Additionally, they have audio capture and motion detection to check for wearer compliance which, when used with software, supports your noise exposure compliance and hearing conservation program.  This means easy analysis of the results with professional reports saving yourself time. Audio files of events and motion can be analysed to determine if any of the data is erroneous, which can be quickly and easily removed from exposure data, giving confidence in the results. And should you have any spurious noise readings; these can be selected and excluded from exposure results.

So, what about the future? No doubt we will see incremental advances in professional noise monitoring instrumentation that add functionality and further improve ease of use. Perhaps there is a technology-based paradigm shift just around the corner? Standards that lag technology need to play catch up and better still anticipate the direction of travel. For the time being, the safest bet (no pun intended) is to use a professional noise measuring instrument for your quantitative risk assessment.

References

1.       IEC 61672:2002 Part 1 – Electroacoustics – Sound level meters – Part 1: Specifications

2.       IEC 61252:1993+AMD1:2000+AMD2:2017 - Noise Dosimeters-Specifications for personal sound exposure meters (under review 2022)

3.       ANSI S1.4- 2014 (R2017) Specification for sound level meters

4.       ANSI S1.43-1997 (R2007) Specifications for Integrating-Averaging Sound Level Meters

5.       ANSI/ASA S1.25-1991 (R2017) Specification for Personal Noise Dosimeters

6.       ISO9612:2009 Acoustics — Determination of occupational noise exposure — Engineering method

7.       OSHA 29CFR 1910.95 - Standard on Occupational Noise Exposure

8.       IEC61260 Ed. 1.0 (2014) Electroacoustics – Octave-band and fractional-octave-band filters & ANSI S1.11-2004 (R2009)

9.       https://www.wikiaudio.org/equal-loudness-contour/

10.     ISO 226:2003 Acoustics — Normal equal-loudness-level contours

11.     https://www.cdc.gov/niosh/topics/noise/app.html

12.     Regulation (EU) 2016/425 (implemented in the UK, The Personal Protective Equipment (Enforcement) Regulations 2018)