Sound is measured using a combination of instruments and units that capture its two most important physical qualities: how loud it is and how high or low its pitch is. The core unit of loudness is the decibel (dB), a logarithmic scale that compresses the enormous range of pressures your ear can detect into manageable numbers. Pitch is measured in hertz (Hz), counting how many pressure wave cycles pass a point each second. The workhorse instrument for most practical situations is the sound level meter, but the full toolkit extends from smartphone apps to laser-based vibration sensors and massive infrasound monitoring networks, each designed for a different slice of the acoustic world.
The Decibel Scale and Why It Is Logarithmic
The quietest sound a healthy human ear can detect and the loudest sound that causes immediate pain differ in pressure by a factor of roughly ten million. Working with numbers that span such a wide range in a linear way would be impractical, so acousticians use the decibel, a unit based on the logarithm of the ratio between a measured pressure and a reference pressure. That reference is set at 20 micropascals, which represents the approximate threshold of human hearing at 1,000 Hz. On this scale, a whisper in a quiet room registers around 30 dB, normal conversation sits near 60 dB, a lawnmower at close range reaches about 90 dB, and a rock concert can push past 110 dB.
Because the scale is logarithmic, every increase of 10 dB roughly corresponds to a doubling of perceived loudness. That means 80 dB does not sound “a little louder” than 70 dB; it sounds about twice as loud. This relationship between the physical measurement and the human experience is one reason the decibel scale is so useful: it maps onto perception more naturally than a straight pressure reading would. It also explains why a difference of just 3 dB, which represents a doubling of acoustic energy, is barely noticeable to most listeners. The gap between “annoying” and “dangerous” is surprisingly small in dB terms.
How Frequency Is Measured and Why It Matters
Frequency, measured in hertz, describes the pitch of a sound. A low rumble of thunder might sit around 40 Hz, while the highest notes on a piano hover near 4,000 Hz. Humans generally hear frequencies between about 20 Hz and 20,000 Hz, though the upper end drops with age. Many sounds we care about, from machinery noise to speech, contain energy spread across a wide range of frequencies rather than a single tone, which is where frequency analysis comes in.
Acousticians rarely look at every individual frequency. Instead, they group sound energy into bands. Historically, these bands were based on the octave, a doubling of frequency, and further divided into sub-bands like one-third octave bands. This approach was first standardized in the 1950s. Because ten of those one-third octave bands span a range close to one decade of frequency, modern standards have shifted to decidecade bands, which divide each decade into ten equal parts on a logarithmic scale. The shift simplified international standards while keeping results practically identical to the older system.1The Journal of the Acoustical Society of America. Fractional octave and fractional decade frequency bands in acoustics: Historical review and recommendations You will encounter both terminologies depending on the age of the equipment or standard being used, but they describe the same basic idea: slicing the spectrum into chunks that are meaningful for human hearing or engineering analysis.
Sound Level Meters and Microphones
The most common instrument for measuring sound is the sound level meter, or SLM. In its simplest form, an SLM consists of a microphone that converts pressure fluctuations into an electrical signal, a set of electronic filters, and a display that reports the result in decibels. Professional-grade meters conform to international standards (IEC 61672 is the current one), which define two accuracy classes. Class 1 instruments are required for precision work such as regulatory compliance and laboratory measurements, while Class 2 instruments are acceptable for general-purpose surveys.
The microphone inside an SLM is typically a condenser type, which uses a thin diaphragm whose movement relative to a fixed backplate generates a voltage proportional to sound pressure. Condenser microphones offer wide frequency response and good sensitivity, making them the default for acoustic measurement. For harsher environments, like engine test cells or outdoor monitoring stations exposed to weather, specialized microphones with protective windscreens, heated preamplifiers, or high dynamic range capsules are used instead.
Most meters apply one or more frequency-weighting curves to the raw signal before displaying a number. The most common is the A-weighting curve (yielding readings in “dBA”), which rolls off low and very high frequencies to mimic the reduced sensitivity of the human ear at those extremes. A-weighted levels are used in most occupational noise regulations and environmental noise ordinances. C-weighting (dBC) keeps more low-frequency content and is used when you want to assess the full physical energy, for example when evaluating hearing protection or measuring bass-heavy noise from industrial equipment.
Smartphone Apps as Measurement Tools
Given that most people carry a microphone-equipped computer in their pocket, it is no surprise that dozens of smartphone apps claim to measure sound levels. The question is whether they are accurate enough to be useful. A study evaluating these apps found that iOS-based sound measurement apps could be within about 2 dB of reference-grade sound level meters when measuring steady-state noise, though performance varied considerably depending on the phone model and the specific app.2PubMed Central. So How Good are These Smartphone Sound Measurement Apps?
A 2 dB margin sounds impressive, but there are caveats. Smartphone microphones are not calibrated at the factory for acoustic measurement, their frequency response varies between models and even between individual units, and external factors like a phone case or your hand placement over the microphone port can skew results. For a rough check of whether your office is uncomfortably loud or whether a concert is approaching dangerous levels, an app is a reasonable first pass. For anything with legal, regulatory, or clinical stakes, you need a proper sound level meter with a current calibration certificate.
Sound Intensity Probes
A sound level meter measures sound pressure at a single point, but it cannot tell you which direction the sound is coming from or how much acoustic power is actually flowing through space. That is the job of a sound intensity probe, which measures both the pressure and the particle velocity of the sound wave to determine the magnitude and direction of energy flow.
The most widely used design is the two-microphone pressure-pressure probe. It uses a pair of closely spaced condenser microphones; the difference in pressure between the two gives an approximation of particle velocity, and the average of the two gives pressure. Together, these yield the sound intensity vector. The theory and technology behind these probes have been developed and refined over decades, with particular attention paid to the factors affecting accuracy, such as microphone spacing, phase mismatch between channels, and the frequency range over which the technique is valid.3Journal of Vibration and Acoustics. The Two-Microphone Sound Intensity Probe
More recent work has explored miniaturized and cost-effective versions. One design uses one or two miniature electret microphones in a mechatronic system to determine the sound intensity vector on a plane, making 2D intensity mapping more accessible for smaller budgets and tighter spaces.4PubMed Central. Performance Analysis of Cost-Effective Miniature Microphone Sound Intensity 2D Probe Intensity measurements are especially valuable for identifying which part of a machine is actually producing the noise, since they can pinpoint sources even in the presence of background sound that would confuse a simple pressure measurement.
Acoustic Cameras and Beamforming Arrays
When you need to locate sound sources visually, acoustic cameras combine a video camera with an array of microphones arranged in a geometric pattern. Signal-processing techniques, collectively called beamforming, are applied to the microphone signals to determine where in space each sound is coming from. The result is typically a color-coded map overlaid on the video image, showing hot spots of noise. Acoustic cameras are used in automotive testing, aircraft engine noise studies, building facade surveys, and factory troubleshooting. Different beamforming techniques can be applied to the captured signals, allowing users to visualize noise and distinguish the contributions of multiple sources even when they are emitting at the same time.5INTER-NOISE and NOISE-CON Congress and Conference Proceedings. Influence of emitter’s position in the field of view of the acoustic camera and ground reflection in sound source localization with beamforming
One limitation of purely microphone-based beamformers is that they can struggle in complex acoustic environments with multiple reflections, such as rooms with hard walls. A newer approach pairs a planar microphone array with a 360° camera, using the visual information about source location and orientation to steer and refine the beamformer, improving performance in reverberant settings.6The Journal of the Acoustical Society of America. A novel audio-visual based beamformer using a 360° camera and planar microphone array system This kind of sensor fusion, combining audio and video data, is becoming more common as both camera and microphone arrays become cheaper.
Laser-Based Measurement Without Contact
Sometimes you cannot place a microphone where the sound is happening. The vibrating surface might be inside a sealed chamber, spinning at high speed, or too hot to touch. Laser Doppler vibrometry offers a way to measure vibrations, and by extension the sound they produce, without any physical contact. The instrument splits a laser beam into two paths: one aimed at the vibrating surface and one kept as a reference. When the reflected beam recombines with the reference beam on a detector, the phase difference between them reveals the velocity and displacement of the surface with extraordinary precision.7Scientific Reports. Non-contact photoacoustic imaging with a silicon photonics-based Laser Doppler Vibrometer
Laser vibrometers are used extensively in loudspeaker design, where engineers need to see exactly how a diaphragm deforms at different frequencies, and in non-destructive testing of materials where internal flaws alter surface vibration patterns. Miniaturized versions built on silicon photonics platforms are bringing costs down and opening the door to integrated sensing systems that could eventually replace some microphone-based measurements entirely.
Measuring Sounds You Cannot Hear
Human hearing covers roughly 20 Hz to 20,000 Hz, but sound does not stop at either boundary. Below 20 Hz lies infrasound, and above 20,000 Hz lies ultrasound. Both require specialized measurement approaches.
Infrasound
Infrasound is produced by volcanic eruptions, large explosions, severe weather, ocean waves, and even wind turbines. The International Monitoring System (IMS), originally built to detect nuclear tests, operates an infrasound network designed for atmospheric pressure fluctuations in the 0.02 Hz to 4 Hz range. Most of the measuring equipment in this network also records fluctuations at even lower frequencies, which researchers have used to study atmospheric gravity waves with periods ranging from a few minutes to 24 hours.8Geophysical Research Letters. Using the International Monitoring System infrasound network to study gravity waves
Standard condenser microphones are not suitable for infrasound because their low-frequency response rolls off well above 0.02 Hz. Infrasound sensors typically use microbarometers, which are sensitive differential pressure transducers designed to respond to very slow pressure changes. Calibrating these sensors at such low frequencies is its own challenge. One solution is the infrasonic pistonphone, a device that uses a motor-driven piston to generate known pressure oscillations. A simple design covering 20 millihertz to 20 Hz has been developed with digital correction for the mechanical distortion inherent in a crank-driven piston, enabling accurate calibration without expensive or complex equipment.9The Journal of the Acoustical Society of America. Infrasonic pistonphone calibration of acoustic transducers
Ultrasound
At the other extreme, medical and industrial ultrasound operates at frequencies from roughly 20,000 Hz into the megahertz range. At these frequencies the quantity of interest is often acoustic power rather than pressure, because the energy delivered by an ultrasound beam determines its therapeutic or imaging effect. The standard method for measuring ultrasound power is the radiation force balance, which works by directing the beam at a target (either absorbing or reflecting) suspended from a sensitive scale. The force the beam exerts on the target is directly proportional to the acoustic power.
The radiation force balance is widely used for calibrating medical ultrasound equipment. For highly focused beams used in therapeutic applications, proper alignment between the beam and the target is critical to accurate measurements.10PubMed Central. Image-Guided Measurement of Radiation Force Induced by Focused Ultrasound Beams Work on high-intensity focused ultrasound (HIFU) systems has demonstrated that radiation force balances with absorbing targets can measure acoustic power up to 500 watts with good linearity.11PubMed. Acoustic power measurement of high intensity focused ultrasound in medicine based on radiation force That upper end represents serious power, enough to heat and destroy tissue in surgical applications, which is exactly why accurate measurement matters.
Loudness Is Not the Same as Sound Pressure
One of the most persistent misconceptions about sound measurement is that “louder” always means “more decibels.” In physical terms, a 70 dB tone at 1,000 Hz and a 70 dB tone at 100 Hz have the same sound pressure level. But they do not sound equally loud. Your ear is much less sensitive at low frequencies, so the 100 Hz tone sounds considerably quieter even though the meter reads the same number. This disconnect between what the meter says and what the listener hears is the core problem of psychoacoustic measurement.
The relationship between frequency and perceived loudness is captured by equal-loudness contours, which map out the sound pressure level needed at each frequency for a listener to judge the sound as equally loud as a reference tone at 1,000 Hz. These contours are defined in the international standard ISO 226. However, research has found that the contours specified in that standard involve substantial errors, particularly for frequencies below 1,000 Hz.12The Journal of the Acoustical Society of America. Estimation of the new equal-loudness level contours Revised contours have been proposed using improved loudness models, and the standard has been updated over the years, but the fundamental challenge remains: translating a physical measurement into a number that predicts human experience is inherently messy.
This is also why the A-weighting curve mentioned earlier exists. A-weighting is essentially a simplified, single-curve approximation of the equal-loudness contours at moderate levels. It works reasonably well for everyday noise assessment, but it underestimates the impact of low-frequency noise, which is why communities near wind farms or industrial bass-heavy sources sometimes report annoyance at levels that look acceptable on an A-weighted meter. C-weighting or unweighted measurements are often more informative in those cases.
Before Electronic Instruments
It is easy to assume that measuring sound has always required electronics, but a significant body of acoustic pressure measurements was made before electronic amplification existed. The key instrument was the phonodeik, developed by physicist Dayton C. Miller in the first decade of the twentieth century. The phonodeik used a thin diaphragm connected by a silk thread to a tiny mirror. Sound pressure moved the diaphragm, which pulled the thread, which rotated the mirror. A beam of light reflected off the mirror traced the pressure waveform onto a moving photographic film, producing a visual record of the sound wave.13The Journal of the Acoustical Society of America. The phonodeik: Measuring sound pressure before electroacoustic transducers
Miller used the phonodeik to study the waveforms of musical instruments, speech, and various noise sources, producing detailed photographs of sound waves decades before oscilloscopes were available. The device was purely mechanical and optical, with no electrical components involved. By today’s standards it was crude, limited in bandwidth and sensitivity, but it established the principle that sound is a measurable physical quantity with a definite waveform, not just a subjective sensation. That conceptual leap, from treating sound as something you hear to something you can record, quantify, and compare, laid the foundation for every instrument described above.
Calibration and Traceability
Even the finest microphone or the most advanced beamforming array is only as trustworthy as its calibration. Acoustic instruments are calibrated by exposing them to a known sound and checking whether their reading matches. For everyday sound level meters, a handheld acoustic calibrator produces a tone at a fixed level, typically 94 dB or 114 dB at 250 Hz or 1,000 Hz, and you verify that the meter reads correctly. Periodic calibration (often annually at a certified lab, plus a field check before each measurement session) is a legal requirement for many regulatory applications.
Calibration gets more challenging at extreme frequencies. At infrasonic frequencies, the pistonphone described earlier is one of the few practical tools, generating known pressure oscillations as low as 20 millihertz.9The Journal of the Acoustical Society of America. Infrasonic pistonphone calibration of acoustic transducers At ultrasonic frequencies, hydrophone calibrations are typically performed in temperature-controlled water tanks using known reference transducers, with traceability back to national measurement institutes. The entire chain from field instrument to national standard must be unbroken; without it, a measurement is just a number on a screen with no guaranteed relationship to reality.
For microphone arrays and intensity probes, phase calibration between channels is just as important as amplitude calibration. A tiny phase mismatch between two microphones in an intensity probe can produce large errors in the calculated intensity, especially at low frequencies where the pressure gradient is small. Manufacturers specify residual phase-mismatch tolerances, and users need to verify those tolerances periodically, a step that gets skipped more often than it should in field practice.