Skip to content
Test Utilities

SPL meter

A-, C-, or Z-weighted sound-level meter with Fast / Slow integration and one-shot calibration.

Live broadband SPL from your mic, computed in the frequency domain with IEC 61672-1 A / C weighting curves and the standard Fast (125 ms) and Slow (1 s) time constants. Calibrate against a 94 / 114 dB acoustic calibrator and the readout becomes absolute SPL; any other reference applies an offset but the meter stays flagged relative-only. The offset persists per browser and is invalidated by any mic or OS input-gain change. With a UMIK-1 or iMM-6 measurement mic, expect ±1 dB. With a laptop mic, treat absolute level as approximate but trends as meaningful. Audio never leaves your browser.

mic off — press start to measure
Weighting
Time
View
input

Grant mic access to see your input devices.

Sets the precision estimate. Doesn't shift the reading.

calibration

How to get absolute SPL

The browser can't read absolute SPL on its own — your mic's sensitivity is unknown to it. Anchor the meter to a known reference and hit Capture.

  1. Acoustic calibrator (B&K 4231, miniDSP cal, etc.) — drop the mic in the coupler, pick 94 (or 114) dB, capture. This is the only path that marks the meter calibrated.
  2. Second meter / phone SPL app — hold it next to your mic with a steady sound playing, type its reading in, capture. The offset is applied, but the meter stays flagged relative only — you've inherited that device's error plus placement mismatch.
Calibrator presets
current offset: none

Tip · Keep the source steady for the full 750 ms capture. Calibrators emit a 1 kHz tone where A, C and Z all agree, so weighting doesn't matter for path 1; for a second-meter reference, match this meter's weighting to the other device's.

how to use
  • A-weighting mimics how the ear hears mid frequencies — use it for general loudness, OSHA dose, and most environmental measurements.
  • C-weighting keeps low-end energy in the count — use it for live-sound peaks, subs, and PA check-outs.
  • Fast (125 ms) tracks transients; Slow (1 s) averages out chatter for steady-state readings.
  • The offset persists per browser, but it is only valid for the exact mic and input gain it was captured with. Any OS / audio-interface input-gain change, a different mic, or a different USB port silently invalidates it — recapture.
  • Lmax (5 s) is the running maximum of the time-weighted level; it holds for five seconds after the last new high, then decays at 6 dB/s.
  • Audio never leaves your browser. Mic → analyser only — nothing recorded, nothing uploaded.

Need this math wired into a real system? We do that for a living.

Start a conversation →