Using a smartphone sound level meter application in a field noise measurement. Source: NIOSH / Wikimedia Commons, public domain.
When Canadian municipal noise bylaws specify a decibel limit — for example, 55 dB(A) during daytime hours in a residential zone — the "(A)" carries important technical meaning. It refers to A-weighting, a frequency filter applied to the raw sound signal before the level is calculated. Understanding why this filter exists, and how it changes the reading, is essential for interpreting noise measurement results correctly.
Why Human Hearing Is Not Flat
Human hearing does not respond equally to all frequencies. The ear is most sensitive in the range of roughly 1,000 to 4,000 Hz — frequencies that overlap with speech and many mechanical tones. At lower frequencies (below 500 Hz) and very high frequencies (above 8,000 Hz), the ear is considerably less sensitive. A 40 Hz hum from a mechanical ventilator and a 2,000 Hz tone from a power tool might produce the same physical sound pressure level, but the 2,000 Hz tone will be perceived as much louder by most listeners.
Unweighted decibel measurements (sometimes called dB or dBZ) capture all frequencies equally. A meter reading 65 dB (unweighted) near a diesel generator running at idle tells you about the physical energy of the sound, but not about how a nearby resident is likely to perceive it. If most of that energy is concentrated at low frequencies the ear is less sensitive to, the actual perceived annoyance may be lower than the raw number suggests.
What A-Weighting Does
A-weighting applies a standardized frequency-response curve to the incoming signal before calculating the level. The curve amplifies the range around 1,000–4,000 Hz and attenuates low frequencies and very high frequencies, approximating the average human hearing response at moderate sound levels.
The resulting number — expressed as dB(A) — correlates much better with perceived loudness and, by extension, with the likelihood that a sound will cause annoyance, sleep disturbance, or hearing impairment over sustained exposure. This is why international standards including ISO 1996 (environmental noise) and occupational health standards such as ACGIH TLVs specify dB(A) for their limits.
The A-weighting curve is defined in IEC 61672-1 (the international standard for sound level meters) and is reproduced in Canadian-referenced standards including ANSI S1.4. Sound level meters sold for environmental noise measurement are expected to implement A-weighting within specified tolerance limits.
Other Weighting Curves
A-weighting is not the only option. C-weighting is a flatter curve that attenuates far less at low frequencies, making it useful for assessing peak sound pressure levels and for evaluating sources with strong low-frequency content, such as bass from amplified music or distant industrial machinery. Some bylaws require both an A-weighted and a C-weighted reading to assess bass-heavy sources. Z-weighting (flat response) is used in specialized acoustic research contexts.
Canadian environmental noise bylaws almost universally specify dB(A) as the regulatory metric for residential and commercial noise limits. C-weighting appears in some industrial contexts and in occupational noise exposure standards, but rarely as the primary measure in a municipal bylaw.
Time Weighting: Fast, Slow, and Impulse
A sound level meter's A-weighting filter addresses which frequencies are included; a separate setting — the time constant — determines how quickly the meter responds to changes in level. The two standard settings are:
- Fast (F): 125 ms time constant. Responds quickly to level changes, tracks fluctuating noise more closely in real time.
- Slow (S): 1 second time constant. Averages over a longer period, producing a more stable reading for sustained or slowly varying noise.
Vancouver's Noise Control Bylaw explicitly requires measurements to be made with the slow time-weighting setting for enforcement purposes. This is a common requirement in Canadian bylaws, as it prevents temporary peaks (a car door slamming, a brief shout) from triggering a reading that does not reflect sustained noise conditions.
Impulse weighting (I) uses a very fast rise time and is used to assess impulsive sounds such as pile-driving or pyrotechnics. Some bylaws include separate limits or provisions for impulsive noise sources.
Equivalent Continuous Level (Leq)
A single spot measurement with a sound level meter captures the level at one moment. For noise that varies over time — street traffic, outdoor events, HVAC cycling — a more meaningful metric is the equivalent continuous sound level, denoted Leq (or Leq). This is the steady level that would contain the same total acoustic energy as the actual fluctuating noise over the measurement period.
Most enforcement-grade sound level meters can calculate Leq over a measurement interval (commonly 1–15 minutes). Municipal inspectors assessing a sustained mechanical noise source will typically record an Leq reading rather than a peak or instantaneous level. Ontario's NPC-300 procedure, for instance, specifies Leq measured over defined intervals as the basis for compliance assessment.
Handheld Meters: Type 1 vs Type 2
IEC 61672-1 defines two accuracy classes for sound level meters:
| Class | Accuracy | Typical Use |
|---|---|---|
| Type 1 (Class 1) | ±1 dB (reference frequency) | Laboratory, legal disputes, precision environmental surveys |
| Type 2 (Class 2) | ±1.5 dB (reference frequency) | General field surveys, bylaw enforcement, community monitoring |
Vancouver's bylaw explicitly accepts either Type 1 or Type 2 instruments for enforcement measurements. Other Canadian bylaws typically do not specify a class requirement in the bylaw text itself, but enforcement officers generally use instruments meeting at least Type 2 specifications to ensure measurement credibility if readings are challenged.
Consumer smartphone applications can estimate sound levels but do not meet IEC 61672-1 requirements. Readings from phone apps are not considered credible evidence in formal noise complaint proceedings, though they can serve as a useful preliminary indicator of whether a problem warrants further investigation.
A modern handheld sound level meter with a 120 dB measurement range. Source: Beccandcal / Wikimedia Commons, CC BY-SA 3.0.
Calibration Requirements
A sound level meter is only as accurate as its calibration. IEC 61672-1 requires that instruments be calibrated using an acoustic calibrator (a pistonphone or similar device) before and after each measurement session. Without a calibration record, the credibility of any measurement taken can be challenged. Municipal enforcement departments typically maintain calibration logs for their instruments and may be required to produce them if a noise violation is contested.
Calibrators themselves must be traceable to national standards. In Canada, this means traceability to the National Research Council Canada's (NRC) measurement standards.
What Affects a Field Reading
Several site conditions can affect the accuracy and representativeness of a field noise measurement:
- Wind: Wind striking the microphone generates turbulence noise that inflates readings. Windscreens (foam or hairy ball covers) are required in outdoor measurements above about 1–2 m/s wind speed. Readings taken in gusty conditions without windscreens are unreliable.
- Microphone position: Measurements near reflective surfaces (walls, fences) will be elevated by reflected sound. Standard practice is to measure at least 0.5–1 m away from any reflecting surface, or to apply a correction factor.
- Background noise: If the background level is within 3–10 dB of the source being measured, a correction is needed to isolate the source contribution. If the background is less than 3 dB below the total, it is not possible to reliably separate the source from the background using simple A-weighted level measurements.
- Measurement height: Most bylaws specify measurements at a standard height (commonly 1.5 m) above ground level at the property line of the receiver.