Accuracy and Limitations

This page explains what affects the accuracy of every tool on SoundDBMeter.com, what commonly causes unexpected results, and what the tools fundamentally cannot determine regardless of conditions. It is organised by tool category because the accuracy profile differs significantly between the microphone-based measurement tools, the audio output tool, and the pure calculation tools.

Reading this page before using the tools will help you set appropriate expectations. Reading it after will help you interpret results honestly.

Written and maintained by Addito, founder of SoundDBMeter.com.

Part 1 — Microphone-Based Measurement Tools

This section covers the Online Decibel Meter, Background Noise Test, Frequency Analyzer, and Microphone Test.

These tools capture audio from your device microphone and estimate sound pressure levels or frequency content using the Web Audio API. All results are approximations — not calibrated sound pressure level measurements.

Variable 1 — Microphone Hardware Quality

Consumer microphone quality varies enormously and is the single largest source of measurement variability across devices.

Frequency response. Professional acoustic measurement microphones (used in certified Class 1 and Class 2 sound level meters) are designed to have a flat frequency response across the full audible range. Consumer microphones — including built-in laptop, phone, and tablet microphones — are optimised for speech capture (approximately 300 Hz to 3,400 Hz) and typically roll off significantly at low frequencies. Below approximately 100–150 Hz, most built-in consumer microphones underperform, meaning low-frequency noise sources will be measured as less intense than they actually are.

Microphone sensitivity. Sensitivity varies by manufacturer, model, and even by individual unit. Two phones of the same model may produce different dB readings in the same environment.

Recommendation: For more consistent results, use an external USB microphone with a documented flat frequency response. For precise measurement, use a certified Class 1 or Class 2 sound level meter with a calibrated measurement microphone.

Variable 2 — Operating System Audio Processing

This is the most frequently overlooked accuracy variable in browser-based measurement and one that even professional acousticians may not be aware of.

Modern operating systems apply automatic audio processing to microphone input before it reaches any application — including the browser. This processing typically includes:

  • Automatic Gain Control (AGC) — continuously adjusts microphone gain to maintain a consistent signal level. In a quiet room, AGC amplifies the signal; in a loud environment, it reduces it. This directly counteracts the measurement of actual sound level variation.
  • Noise suppression — attempts to remove background noise from the signal. When activated, it reduces the measured level of consistent broadband noise, making a noisy environment appear quieter than it is.
  • Echo cancellation — removes acoustic feedback, which can alter the signal for sounds that match room resonance characteristics.

These processes are applied at the OS level and cannot be fully disabled or controlled by a browser application. Their effect on measurement accuracy varies by device, OS version, and audio driver settings.

Practical implication: Two devices in the same acoustic environment may produce different dB readings because their OS audio processing applies different gain adjustments and noise suppression settings.

Variable 3 — Browser Web Audio API Implementation

Different browsers implement the Web Audio API with different performance characteristics:

  • Google Chrome — most complete and consistent Web Audio API implementation. Recommended for all tools on this site.
  • Firefox — works in most cases but may differ in sampling rate handling and RMS calculation behaviour.
  • Safari — partial Web Audio API support; some features may behave differently, particularly on older macOS versions.
  • In-app browsers (Instagram, Facebook, TikTok, etc.) — have restricted microphone access and frequently produce unreliable or failed measurements.

Variable 4 — Device Performance

Low-powered devices — older phones, budget tablets, underpowered laptops — may experience processing delays in the Web Audio API that introduce gaps in the measurement cycle. This can result in missed peaks or incorrectly averaged readings.

For the Frequency Analyzer specifically, low-powered devices may produce a lower FFT resolution, reducing the precision of frequency band identification.

Variable 5 — Environmental Measurement Conditions

Background noise. All ambient sound in the measurement environment is captured by the microphone. If you are measuring a specific sound source — a machine, a voice, a speaker — background noise from HVAC systems, traffic, or other sources will be included in the reading. The Online Decibel Meter measures all sound reaching the microphone, not a specific sound source.

Distance and directivity. Sound pressure level decreases with distance from the source. A reading taken one metre from a noise source will be approximately 6 dB higher than a reading taken two metres away (in free field conditions). Microphone directivity — the pattern of sensitivity around the microphone — also affects results when sound is not arriving directly in front of the microphone.

Room acoustics. Reflections, standing waves, and room resonances affect the sound field at the measurement position. A reading in a reverberant room (hard surfaces, small space) will typically be higher than a reading in an acoustically treated space at the same distance from the same source.

Variable 6 — Calibration

Browser-based tools cannot perform automatic acoustic calibration. The Online Decibel Meter provides a manual Calibration Offset control that allows you to align its reading with a reference instrument. This is the recommended approach when comparative accuracy is needed:

  1. Place your device microphone and a reference Class 2 sound level meter at the same location
  2. Note the difference between the two readings at a consistent sound level
  3. Enter the offset into the Calibration Offset field

This does not make the browser tool a certified instrument — it provides a rough correction factor for that specific device in that specific environment.

What Microphone Tools Cannot Determine

Regardless of conditions, the following are outside the scope of browser-based microphone measurement:

Frequency-weighted measurements (dBA, dBC). Consumer microphones do not apply standardised frequency weighting filters. The readings produced by the Online Decibel Meter are unweighted amplitude estimates. True dBA or dBC measurements require a microphone with an IEC-standardised weighting filter.

Regulatory compliance documentation. OSHA and NIOSH compliance monitoring requires certified Class 1 or Class 2 sound level meters with traceable calibration. Browser-based tools cannot produce documentation acceptable for compliance purposes.

Impulse or peak sound levels. The RMS-based measurement used by these tools averages amplitude over a time window. Instantaneous peak sound levels from impact noise (hammer blows, gunshots, explosions) require a peak-hold circuit not available in browser-based processing.

Accurate low-frequency measurement. As documented in Variable 1, consumer microphone low-frequency roll-off means noise sources with significant energy below approximately 100–150 Hz will be underestimated.

Part 2 — Tone Generator

The Tone Generator produces audio output using the Web Audio API’s OscillatorNode, which generates frequencies with negligible error (sub-cent level) in all modern browsers.

What affects the perceived accuracy of Tone Generator output:

  • Speaker and headphone frequency response — the OscillatorNode produces the correct frequency, but your speakers or headphones must be able to reproduce it. Very low frequencies (below approximately 80–100 Hz) will not reproduce accurately through laptop speakers or small phone speakers. Use headphones for full-range frequency testing.
  • Volume level — very low amplitude settings may make tones inaudible or intermittent on some devices.

The Tone Generator is frequency-accurate at the source. The limitation is in the audio reproduction hardware, not the synthesis.

Part 3 — Pure Calculation Tools

The SPL Converter, Volume Level Comparator, and Noise Exposure Calculator perform mathematical calculations on typed numeric inputs. Their outputs are mathematically exact within the formula applied.

The only accuracy consideration for these tools is the quality of the input values:

  • If you enter an estimated dB level from a non-certified source — including the Online Decibel Meter — the calculated result reflects the accuracy of that estimate
  • If you enter a precisely measured, calibrated dB value from a certified instrument, the calculated result will be correspondingly accurate
  • The formulas used are the standard acoustic formulas documented in the Testing Methodology page

These tools do not introduce measurement error — they compute correctly from whatever input they receive.

How to Get the Most Reliable Results

For microphone-based tools:

  • Use Google Chrome on a desktop or laptop
  • Use an external microphone with a flat frequency response where possible
  • Check OS settings for noise suppression and AGC and disable where possible
  • Measure in the quietest representative environment
  • Use the Calibration Offset feature when you have a reference instrument available
  • Take multiple readings and note variation rather than relying on a single reading

For the Tone Generator:

  • Use wired headphones for the most accurate frequency reproduction, especially at low frequencies

For calculation tools:

  • Use calibrated input values where accuracy matters for a specific purpose

Related Pages

This Accuracy and Limitations page is written and maintained by Addito, founder of SoundDBMeter.com.

Last updated: June 2026.

Scroll to Top