How to use this spectrum analyzer online
Press Start, allow the microphone and make a sound. Peaks show the frequencies in it; the readout above the graph names the strongest one in Hz and as a musical note.
- Press Start and choose Allow. This spectrum analyzer online uses your mic live; nothing is recorded.
- Bring the sound source close to the microphone: a voice, an instrument, a fan or a speaker playing a tone.
- Read the peak. The line “Peak 1,000.2 Hz · −18.4 dBFS · B5 +22¢” names the strongest frequency. The table lists the five strongest peaks.
- Freeze (F or Space) to study one moment. Hover or tap the graph to read any frequency.
- Save a reference (R) before you change something, then compare the live trace with the dashed line.
- Export a PNG picture or a CSV file of the spectrum.
Other keys: S starts and stops, P cycles peak hold, L switches log and linear, B switches line and bars, C clears peaks.
How the audio spectrum analyzer works
The audio spectrum analyzer takes a slice of sound from the microphone and runs a fast Fourier transform (FFT). The FFT splits the slice into frequency bins and measures how much energy sits in each one.
The FFT size sets the trade-off. At 48 kHz, 8,192 points give bins 5.86 Hz wide from a 171 ms slice; 32,768 points give 1.46 Hz bins but need 683 ms of sound, so fast changes blur. For detail below 100 Hz, use 16,384 or more.
A log frequency axis gives each octave the same width, which matches how we hear; a linear axis spreads out the highs. Levels are calibrated so a full-scale sine wave reads 0 dBFS. Voice processing (echo cancellation, noise suppression and auto gain) is switched off where the browser allows.
It works as a sound spectrum analyzer online or a simple online frequency analyzer, but it measures dBFS, not sound pressure, and every microphone colors the result.
Reading the spectrum
| What you see | What it usually means | What to do |
|---|---|---|
| One narrow spike | A pure tone, a whistle, or an electronic whine | Read its frequency; coil whine often moves with load |
| Spikes at 50 or 60 Hz and multiples | Mains hum from a ground loop or a cable | See the hum steps below |
| Evenly spaced peaks | A voice or an instrument; the lowest peak is usually the note | Use the online tuner for pitch |
| A broad hump | Fan noise, air or water flow | Compare before and after with a reference |
| A cliff near 16 kHz | A lossy file, or the mic’s own limit | Normal for MP3s and phone mics |
| Bass peaks during a sweep | Room modes boosting some bass notes | Move the speaker or the listening spot |
The strongest peak can be a harmonic rather than the note you hear, which is why small markers point at the harmonics of the strongest peak.
Finding hum, whine and room problems
Hum: the hum finder reports “Mains hum at 50 Hz” or “60 Hz” when it sees two or more harmonics of the power line frequency (50 Hz in most of Europe, Asia and Africa, 60 Hz in North America). Unplug devices one at a time, try a different outlet for everything in the chain, and use balanced cables or a ground-loop isolator if the hum stays.
Whine: a high, narrow spike that changes when a computer or charger works harder is usually coil whine. Note the frequency and whether it follows the device when you move it.
Rooms: play a slow sweep from the tone generator on your speakers and watch the peak hold trace for bass bumps. The subwoofer test has bass tones for the same check. Save a reference, move the speaker, and compare.
If no peaks show up at all, run the mic test to check that the right microphone is selected and not muted.
Spectrum analyzer vs spectrogram
A spectrum analyzer shows one moment: level against frequency, right now. A spectrogram shows frequency over time as a scrolling picture, which is better for speech, birdsong, or a noise that comes and goes. To analyze an audio file such as an MP3 instead of the microphone, use the spectrogram, which can open files.