The Short Answer
A well-built online decibel meter using a phone or laptop microphone can give you useful, repeatable estimates for comparison and education when set up carefully, but the error is device-dependent and can be several decibels or more — different devices routinely disagree by 5–10 dB or more on the same sound, and no universal accuracy can be promised without calibration against a trusted reference. It is not a certified measurement instrument, and readings are not suitable for legal, medical, occupational, or regulatory decisions. Results outside the mid-range (very quiet or very loud) are less reliable.
That range of usefulness covers most everyday questions: is the room 40 or 55 dB? Is the street 65 or 75 dB? Is the speaker dangerously loud? For those, a phone-based online decibel meter can be a practical screening tool once you have calibrated it and use averages over 30–60 s. For legal disputes, workplace compliance, medical assessment, or hearing-damage claims, it is not — that requires a calibrated sound level meter meeting a standard such as IEC 61672. See Validation and Methodology for what we have (and have not) verified.
Why Microphone Hardware Limits Accuracy
Consumer microphones — especially the tiny ones inside phones and laptops — are engineered for speech and video calls, not for flat, predictable frequency response. Their characteristics vary from model to model and even between units of the same model:
- Frequency response: a mic that boosts mid-frequencies (voice range) and cuts lows will over-report some sounds and under-report others.
- Sensitivity: each device maps its own electrical signal level to a dB value, which is why two phones measuring the same sound differ.
- Dynamic range: very quiet sounds fall below the mic's noise floor, and very loud sounds clip — readings above ~100 dB from a phone should be treated with extra skepticism.
- Directionality: many mics are slightly directional; turning the device can change the reading by several dB.
The phone decibel meter guide covers device-specific behavior in more depth.
Automatic Gain Control, Noise Suppression & Echo Cancellation
These three features are the hidden killers of mobile measurement accuracy. They are designed to make calls sound good, and they quietly modify the audio signal before any meter software can analyze it:
- Automatic gain control (AGC): turns the microphone's sensitivity up in quiet rooms and down in loud ones — precisely the opposite of what a measuring instrument should do.
- Noise suppression: filters out "background" sounds, which is exactly the sound you may be trying to measure.
- Echo cancellation: removes sounds it considers replayed audio, again altering the signal.
Most browsers let websites request these to be disabled when capturing audio, and this tool requests
echoCancellation: false, noiseSuppression: false, and
autoGainControl: false. However, some devices and browsers ignore these requests — check the
"Device & Microphone Information" panel in the meter to see what your device actually applied.
Room Acoustics, Distance and Orientation
Even with a perfect microphone, the environment changes the reading:
- Reflections: in a hard-walled room, sound bounces off surfaces and raises the level at the microphone compared with outdoors.
- Distance: levels fall roughly 6 dB per doubling of distance outdoors (less indoors). "The street is 70 dB" means nothing without a distance.
- Orientation: pointing the phone away from the source, or covering the mic with a case or finger, easily costs 3–10 dB.
- Body shielding: holding the phone against your body blocks and reflects sound irregularly.
These factors affect professional meters too — which is why standards define exact microphone placement. The how-to guide walks through getting consistent, comparable results.
What Calibration Can and Cannot Do
Calibration applies a fixed offset so that the meter agrees with a known reference level (for example, a 94 dB calibration tone or a separate, trusted meter). This genuinely improves consistency: it corrects for your specific microphone's sensitivity at that level.
What calibration cannot do is fix the microphone's frequency response or dynamic range. A mic that boosts the voice range will still read differently for bass-heavy music than a laboratory mic, even after calibration. Think of calibration as leveling the scale — not replacing the instrument. The calibration guide shows the exact steps.
When You Need a Certified Sound Level Meter
A phone — any phone, with any app — is not a substitute for a certified meter in these situations:
- Workplace noise assessments for regulatory compliance (OSHA, EU directives, and similar).
- Legal evidence for noise complaints, boundary disputes, or construction claims.
- Any measurement that will be challenged or audited.
- Measurements below ~30 dB or above ~100 dB where consumer microphones are least reliable.
Certified sound level meters meet international standards (for example, IEC 61672) and are calibrated against traceable references. If your measurement needs to withstand scrutiny, use one — and have it calibrated by a competent laboratory.
How to Get the Most Accurate Readings Possible
- Disable audio processing in your device settings where possible (AGC, noise suppression, echo cancellation).
- Calibrate against a known reference before an important measurement.
- Measure at a consistent distance and note the distance with every reading.
- Keep the mic unobstructed — no case over the port, no finger, no pocket.
- Point the mic toward the source and hold the phone still.
- Average over time: take the average or Leq over 30–60 seconds rather than trusting a single instant.
- Repeat the measurement from a second position and compare.
For a fully repeatable workflow, follow the step-by-step guide, and for exposure-related questions see the noise exposure guide.