Free online sound meter: the complete guide to measuring decibels in your browser (2026)

An online sound meter turns any device with a microphone into a noise measurement tool. Open a browser, grant mic access, and you see a decibel reading updating in real time. No hardware to buy, no app to download, no account to create.

But most browser-based meters have a critical flaw: they don't disable automatic gain control. The result? Misleading readings that make loud rooms look quiet and quiet rooms look louder than they are.

This guide covers everything: how the technology works, what separates useful tools from decorative ones, accuracy expectations, common mistakes, and when free is genuinely enough vs when you need hardware.

If you want to try it right now: open our sound meter online and click Start.

How a browser-based sound meter works

The browser calls getUserMedia() to access your microphone with specific constraints:

navigator.mediaDevices.getUserMedia({
  audio: {
    echoCancellation: false,
    noiseSuppression: false,
    autoGainControl: false
  }
})

The audio stream connects to a Web Audio API AnalyserNode running a Fast Fourier Transform. Every animation frame (~60/second), the tool:

1. Reads the time-domain audio buffer (2048 samples)
2. Calculates the Root Mean Square (RMS) amplitude
3. Converts RMS to decibels: 20 * log10(rms) + offset
4. Runs FFT for frequency spectrum visualization
5. Updates display with current, peak, average, and frequency data

The math is identical to what professional hardware does. The difference? Calibration and microphone quality.

Why AGC matters (the #1 thing most meters get wrong)

Automatic Gain Control is designed for phone calls - it normalizes audio so quiet voices and loud voices come through at similar levels. With AGC active:

• A 75 dB room gets compressed to look like ~55 dB
• A 35 dB room gets boosted to look like ~45 dB
• Different environments show suspiciously similar readings
• You're measuring the browser's opinion, not the actual room

How to test if your meter disables AGC: Measure next to a running vacuum cleaner. If it reads similar to a quiet room (both around 50-55 dB), AGC is active and compressing your signal. A proper meter should show 65-80 dB near a vacuum.

We tested a dozen browser-based meters against a calibrated reference. Most showed compressed readings because they use default getUserMedia settings. Only tools that explicitly set autoGainControl: false gave honest numbers.

What makes a good free sound meter

After testing multiple tools against a Class 2 reference meter in three environments (quiet bedroom at 32 dB, normal office at 48 dB, busy kitchen at 72 dB), here's what separates useful tools from decorative ones:

Must-have features

1. AGC disabled. Non-negotiable. Without this, readings are meaningless.
2. More than one number. You need current level, time-averaged Leq, peak, and minimum. A single reading is barely useful.
3. Frequency spectrum. A frequency analyzer shows what's making the noise - HVAC hum vs speech vs electronics.
4. Session statistics. Peak, min, average over time. The average (Leq) is what matters for exposure.
5. No friction. No email, no download, no paywall. Works the moment you open it.

Nice-to-have features

• CSV data export for documentation
• Classroom display mode (large visible number)
• Background noise assessment scoring
• Waveform visualization
• Privacy-first design (no server upload)

Accuracy: what to expect from a free decibel meter

An honest breakdown by measurement range:

Key insight: Relative accuracy (comparing spaces with the same device) is consistent within 1-2 dB regardless of calibration. It's absolute accuracy where the gap shows.

The main variable isn't the software - it's your microphone. A 2025 MacBook Pro has a better mic array than a budget 2019 phone. Both answer "is this room at 40 or 80 dB?" - only the precision of the exact number differs.

Free online meter vs professional hardware

The practical question: Does the difference between 72 dB and 75 dB change your decision? If yes, buy hardware. If you just need to know whether you're at 40 or 80, a free tool answers that instantly.

Use cases: what you can measure

Home & living

Checking bedroom noise for sleep (aim under 35 dB). Evaluating a home office for video calls (under 40 dB ideal). Testing if new windows reduced traffic noise. Comparing rooms to find the quietest workspace. Documenting noisy neighbors with timestamped logs.

Office & workplace

Documenting open-plan noise (typically 55-65 dB). Comparing meeting rooms. Justifying workspace changes to facilities management. Quick spot-checks during noise complaints. Manufacturing floors between formal assessments.

Classroom

Teachers project the classroom noise meter on a screen. Students see green/yellow/red feedback and self-regulate without constant verbal reminders. Works on school Chromebooks, no IT approval needed.

Recording & podcasting

Measuring noise floor before podcasting or streaming. Target: below 25 dB for broadcast quality, below 35 for acceptable home recording. The frequency view identifies specific noise sources. Use the background noise test for a structured assessment.

Safety awareness

Quick check if an environment exceeds 85 dB (OSHA 8-hour limit). Not a replacement for formal monitoring, but tells workers if they're obviously in the danger zone. Every 3 dB above 85 halves safe exposure time.

Common noise levels reference

Explore any level interactively: decibel examples tool. For detailed breakdowns: decibel chart explained.

7 common mistakes when measuring noise

1. Measuring too briefly. Under 30 seconds, a single random event (cough, door slam) skews everything. Minimum 60 seconds for a stable average. For characterizing a space, 3-5 minutes.
2. Covering the microphone. Many people hold phones with a finger over the mic port. Know where yours is and keep it unobstructed.
3. Placing device on a vibrating surface. A laptop on a desk picks up typing vibrations, fan resonance, and surface reflections. Hold at ear height or use a stand.
4. Using different devices for comparison. If you measure room A with your phone and room B with your laptop, you're comparing microphones, not rooms. Same device for all comparisons.
5. Ignoring AGC. If readings show 50-55 dB no matter where you go - suspiciously consistent - AGC is probably active and compressing your signal.
6. Reading peaks instead of average. A door slamming hits 80 dB for 0.1 seconds. The room is still at 45 dB. Use the average (Leq) for meaningful assessment.
7. Measuring near reflective surfaces. Sound bouncing off a nearby wall adds 2-6 dB. Measure at least 1 meter from walls for representative readings.

Tips for better readings

1. Use Chrome - most reliable for disabling AGC.
2. Know where your mic port is. Don't block it.
3. Measure at the position where you actually sit or stand.
4. Run for 60+ seconds minimum. 3-5 minutes for thorough assessment.
5. Same device for all comparisons - different mics have different baselines.
6. Close other apps that might use the microphone.
7. Disable system audio enhancements in your OS settings if available.
8. An external USB mic ($30+) improves consistency vs built-in mics.
9. For external noise sources, point the mic toward the sound.

Limitations (being honest)

• No calibration. Absolute values may drift 3-8 dB from truth. Relative comparisons are reliable.
• Microphone limits. Consumer MEMS mics roll off below 100 Hz (miss bass rumble) and clip above 110 dB.
• AGC inconsistency. Chrome disables it reliably. Safari on iOS is inconsistent. Firefox usually cooperates.
• Self-noise floor. Readings below 25-30 dB are unreliable because the mic generates that much noise itself.
• Not for compliance. No regulatory body accepts browser-based measurements. OSHA, EPA, and local ordinances require IEC 61672 certified instruments.
• Frequency limitations. Can't match the full 20 Hz - 20 kHz range of professional meters.

When to buy professional hardware

Spend money on a calibrated meter when:

• Results end up in an OSHA report, legal filing, or building certification
• You need to measure below 25 dB or above 110 dB
• A 3 dB difference changes your action plan
• You need documented measurement uncertainty and traceability
• Long-term unattended monitoring is required

For everything else - and that's 95% of situations where people need a noise reading - a free online tool gives you the answer faster and at zero cost.

FAQ

Bottom line

A free browser-based sound meter won't replace a $2,000 calibrated instrument for OSHA compliance. But it instantly answers the question 95% of people actually have: "Is this room quiet or loud?" That answer - available in 5 seconds with zero cost - is genuinely useful for homes, offices, classrooms, and studios. The technology is sound, the math is correct, and the only real limitation is your microphone hardware.