Online Decibel Meter – Measure Sound Level in Your Browser

No. Without calibration this page shows a weighted digital dBFS reading plus a nominal +100 dB offset, clearly labeled as an estimate for screening, and you can inspect the raw dBFS underneath in Digital Diagnostics. You should rely on calibrated Class 1 or Class 2 hardware for safety, workplace or legal decisions.

iPhone ships with no built-in decibel readout, yet Safari on your iPhone turns this page into a live estimator. You open the site over HTTPS, press Start, grant microphone permission once for this origin, and watch estimated dBA without installing anything. Our iPhone decibel meter guide walks you through positioning and calibration on that exact device.

You do not need to install one for a quick check because this browser page behaves like a free meter on iPhone and Android. You get live dBA, dBC or dBZ alongside current, minimum, Leq, maximum and duration, plus a graph of the last 240 points. Native apps only pull ahead for offline use or background logging while you do other tasks.

Yes, this site itself is the free option and it runs entirely in current Chrome, Edge, Firefox or Safari. You create no account, download nothing, and your audio stays on your device during every run. Just press Start Measuring on the homepage to get an immediate live estimate.

A decibel meter, also called a sound level meter or noise meter, reports sound intensity in decibels from a microphone signal. This online version captures mono audio through getUserMedia and the Web Audio API after you grant permission. You then see that signal shaped by A, C or Z weighting and smoothed by the selected time response.

Sound level meter, noise meter, SPL meter and sound meter all describe the same job of reporting sound in decibels. You will see formal teams use calibrated Class 1 or Class 2 hardware for compliance work, while a browser page gives device-dependent estimates. Here your mono microphone input is processed locally with echo cancellation, noise suppression and automatic gain requested off, though the browser may refuse, and the input is never connected to speakers so you hear no feedback loop.

They give useful screening estimates rather than interchangeable readings, since two phones on the same sound often differ by 5–10 dB before calibration. Phone capsules typically read too high below around 30 dBA, clip near 95–100 dBA, and under-report deep bass. You narrow that gap by calibrating your own phone against a known reference at the same spot.

Think in whole decibels and ignore the decimals, because uncalibrated phone gain alone can shift results by several dB. You get steadier numbers by watching the 30–60 second average or Leq instead of reacting to each jump. Calibrate once per phone, keep the same case and microphone, and never stretch the result into workplace, legal or medical proof. For context, sustained exposure near 85 dBA across 8 hours carries risk that halves in allowed time with every 3 dB increase, while levels under 70 dBA are generally not a concern, so treat phone numbers as screening hints.

Grant the microphone, hold your device at ear height with the mic opening uncovered, stay quiet, and let the reading settle for a few seconds. You should read the average or Leq across 30–60 seconds rather than chasing peaks. Keep distance, weighting choice and Fast or Slow identical whenever you compare two places.

The meter converts microphone sound into an RMS level, derives A, C or Z from the dominant frequency band, then smooths the result with Fast at 125 ms or Slow at 1 s. Leq accumulates the energy average across your run while the spectrum view stays display-only. Your browser performs every step locally through the Web Audio API without uploading audio.

Set your device beside a stable known reference at the same spot, wait for a steady sound, note the difference, and save it as your offset. That correction belongs to one microphone and setup only, so you redo it whenever your case, phone, browser or position changes. Never copy the number across devices because gain and frequency response do not transfer.

This browser meter costs you nothing and asks for no purchase, account or install. Standalone hardware spans budget checkers up through Class 2 instruments costing hundreds and Class 1 instruments costing thousands. You need that calibrated hardware, not a phone estimate, for formal compliance or contract evidence.

Look for calibration profiles, A plus C plus Z weighting, Fast plus Slow response, Leq averaging, and honest export behavior. Here you can tune all of those modes, store up to 20 sessions plus your theme in localStorage, and export CSV or JSON. Calibrated sessions carry dB values while uncalibrated exports deliberately omit dB values to prevent misuse.

Phone hardware captures only part of the low end because small capsules favor speech frequencies and weak bass response is normal. You will see deeper sounds read lower than they feel, especially on A-weighting which discounts lows. Switch to C or Z for bass-heavy music or engines, and bring calibrated low-frequency hardware for serious engineering work.

Your audio is neither uploaded nor stored because all analysis happens inside your browser session. The input path is never routed to speakers, and analytics here count page usage only without ever receiving sound. You can keep working after load with the network muted and watch the live graph continue, which proves the loop stays local.

The idle panel shows -- until you grant microphone access and press Start, which keeps the page honest before data exists. Nothing is simulated or pre-filled while you wait. If you still see the dashes after pressing Start, your browser is still waiting on permission over HTTPS.

dB names the logarithmic sound level unit, while dBA shapes the reading toward human hearing, dBC stays flatter, and dBZ leaves it flat from the dominant band. Fast follows peaks with a 125 ms window and Slow calms them with a 1 s window. You pick A for everyday loudness impressions and C or Z when you want more low-frequency content included.

You can export the current run as CSV or JSON and keep a rolling local history of 20 sessions in this browser without sending anything to a server. dB values appear in those files only while a compatible calibration is active, so uncalibrated exports carry timing and settings without dB numbers. Your saved theme rides along in the same local storage on your own machine.