Sound pressure level (SPL) is measured using a microphone that converts air pressure fluctuations into an electrical signal, which is then expressed in decibels (dB). The decibel scale compares the captured sound pressure against a fixed reference point: 20 micropascals, the faintest sound a healthy human ear can detect. Whether you use a dedicated sound level meter or a calibrated smartphone app, the core process is the same: capture the pressure, apply the right settings, and read the result. Getting an accurate number, though, depends on understanding a few key choices you need to make before and during measurement.
The Math Behind the Decibel Scale
SPL is calculated with this formula: SPL = 20 × log₁₀(p / p₀), where p is the root-mean-square (RMS) sound pressure you’re measuring in pascals, and p₀ is the reference pressure of 20 micropascals (0.00002 Pa). That reference represents the threshold of human hearing, so 0 dB SPL doesn’t mean silence. It means the quietest sound most people can perceive.
Because the formula uses a logarithmic scale, every increase of 6 dB roughly doubles the sound pressure, and every increase of 10 dB sounds roughly twice as loud to your ears. A quiet office sits around 40 dB, normal conversation around 60 dB, and a rock concert can push past 110 dB. For measurements in water, the reference pressure changes to 1 micropascal, which is why underwater dB values look dramatically higher than airborne ones for the same perceived intensity.
Choosing the Right Meter
Dedicated sound level meters are classified under the international standard IEC 61672-1 into two grades. Class 1 meters are precision instruments with a tolerance of ±0.7 dB under reference conditions. They’re used for environmental noise surveys, building acoustics, and legal proceedings where accuracy matters most. Class 2 meters have a tolerance of ±1.0 dB and are designed for general fieldwork like workplace noise assessments and basic environmental checks. For most non-laboratory purposes, a Class 2 meter is sufficient, and many workplace noise regulations explicitly accept them.
If you don’t need a dedicated meter, the NIOSH Sound Level Meter app (available for iOS) is accurate within ±2 dBA and meets Class 2 requirements when paired with a calibrated external microphone. Without that external mic, smartphone apps can still give you a reasonable ballpark for informal checks, but the built-in microphones on phones vary widely in quality and aren’t suitable for compliance measurements or anything with legal implications.
Frequency Weighting: A, C, and Z
Raw sound pressure doesn’t tell you how loud something sounds to a human, because our ears are less sensitive to very low and very high frequencies. Frequency weighting filters adjust the measurement to account for this. You’ll see results reported as dBA, dBC, or dBZ depending on which filter was applied, and picking the right one matters.
A-weighting is the most common. It mimics how we hear at moderate levels, heavily reducing low frequencies (a 63 Hz tone is cut by 26.2 dB, for example) while slightly boosting the 1,000 to 4,000 Hz range where speech lives. Nearly all workplace noise limits and environmental regulations reference A-weighted measurements.
C-weighting is flatter across the spectrum, only cutting the extreme lows and highs. It better represents what we hear at high volumes, when our ears become more sensitive to bass. C-weighting is typically used for measuring peak sound levels. The World Health Organization’s workplace guidelines, for instance, set limits at both 85 dBA for average exposure and 135 dBC for peak levels.
Z-weighting applies no filtering at all. It captures the full, unaltered frequency spectrum and is used when you need to analyze the sound source itself rather than its effect on human hearing, such as testing loudspeaker frequency response in manufacturing.
Time Weighting: Fast, Slow, and Impulse
Time weighting controls how quickly the meter responds to changes in sound level. The Fast setting uses a 125-millisecond time constant, meaning the display updates quickly enough to track most real-world fluctuations. Slow uses a 1-second time constant, which smooths out rapid changes and gives you a more stable reading in environments with lots of short-term variation. Impulse uses a very short 35-millisecond rise time designed to capture brief, sharp sounds like hammering or gunfire.
For general-purpose measurements, Fast is the default choice. Use Slow when you want a steadier number in a fluctuating environment, and Impulse when the sound you care about is a sudden burst.
Leq and Other Averaging Metrics
A single instantaneous reading rarely tells the full story. Most real-world sound environments fluctuate constantly, so professionals rely on time-averaged metrics. The most important is Leq (equivalent continuous sound level), which represents the constant sound level that would deliver the same total sound energy as the actual fluctuating sound over a given period. Think of it as a true energy average rather than a simple numerical average.
Other useful metrics include Lmax and Lmin (the highest and lowest levels during a measurement period) and statistical levels like L10, L50, and L90, which tell you the level exceeded 10%, 50%, or 90% of the time. L90, for example, is often used to describe background noise because it reflects the level present for most of the measurement period. For occupational noise exposure, the key number is an 8-hour time-weighted average (TWA). NIOSH recommends a limit of 85 dBA TWA with a 3 dB exchange rate, meaning every 3 dB increase halves your safe exposure time.
Calibrating Before You Measure
A sound level meter should be calibrated in the field immediately before and after every measurement session. This is done with an acoustic calibrator: a small device that fits over the microphone and generates a known tone, typically 94 dB or 114 dB at 1,000 Hz. You place the calibrator over the mic, wait about 20 seconds for the output to stabilize, and adjust the meter until it reads the expected value. If the before and after calibration readings differ by more than about 0.5 dB, the data collected during that session may be unreliable.
Calibrators come with adapters for different microphone sizes (1 inch, 1/2 inch, 1/4 inch). Make sure yours matches your meter’s microphone, and that both the calibrator and the meter are within their laboratory recertification period, which is usually annual.
Microphone Position and Orientation
Where and how you hold the meter affects your reading. For general environmental noise, position the microphone at the height and location where the sound exposure actually occurs: roughly ear height for a standing person, about 1.2 to 1.5 meters from the ground. Keep the meter at arm’s length or mount it on a tripod to minimize reflections from your body.
Sound level meters typically use omnidirectional microphones, which pick up sound equally from all directions. Most are designed for either free-field use (point the mic directly at the source, 0 degrees) or random-incidence use (point the mic 90 degrees away from the source, toward open space). Your meter’s manual will specify which type of microphone it has. Using the wrong orientation can introduce errors of several decibels at higher frequencies.
Dealing With Wind and Other Interference
Wind is the biggest enemy of outdoor SPL measurements. Even a light breeze creates turbulence around the microphone diaphragm that registers as low-frequency noise, potentially adding 10 to 20 dB of error. A foam windscreen reduces this dramatically, and you should always use one outdoors. The tradeoff is minor: foam windscreens can slightly attenuate high-frequency response above 10 kHz, depending on the density of the foam, but for most A-weighted measurements this effect is negligible.
Other sources of interference include reflections from nearby walls and hard surfaces, electrical noise from machinery, and vibration transmitted through the surface where the meter is mounted. When possible, measure in the center of open spaces rather than near reflective boundaries, and use a vibration-isolating tripod mount rather than setting the meter on a table or ledge that might conduct structural vibrations.
Putting It All Together
A typical measurement session follows a straightforward sequence. First, attach the windscreen and calibrate the meter using your acoustic calibrator. Select A-weighting and Fast time response unless your specific application calls for something else. Position the microphone at the correct height and orientation, away from reflective surfaces and your own body. Start recording, and let the meter run long enough to capture a representative sample. For steady noise sources, 30 seconds to a minute is often enough. For fluctuating environments, you may need several minutes or even hours to get a meaningful Leq. When you’re done, recalibrate and compare to your initial reading to confirm the meter held its accuracy throughout the session.