Free online audio spectrum analyzer. Point it at your microphone, an audio file or a built-in test signal and watch every frequency in real time as a 1/3 octave RTA, an FFT curve or a scrolling spectrogram, with peak hold, averaging, A-weighting and exports.
Use your microphone to see the sound around you, open an audio file, or play a test signal to see how a known sound looks.
Everything runs inside your browser. Microphone sound and audio files never leave your device.
An audio spectrum analyzer splits a sound into its frequencies, the same way a prism splits light into colors. Left is low and deep, right is high and bright, and the height shows how strong each part is.
Switch views at the top of the analyzer. They show the same sound in three ways, each one good for a different job.
A smooth line through thousands of frequency points. Best for finding exact pitches, hum, feedback and narrow peaks. The dashed line holds the highest level reached.
31 bars from 20Hz to 20kHz, each one a third of an octave wide, like a graphic equalizer. Best for judging tonal balance and tuning a room or a PA system with pink noise.
Time runs left to right and pitch runs bottom to top, with stronger sound shown darker. Best for speech, birdsong, sweeps and anything that changes over time.
Once you know these patterns, you can spot them anywhere. The ones with a button can be played straight into the analyzer above.
A single sharp spike at one frequency and nothing else. Test tones and tuning forks look like this.
A main spike plus a row of smaller ones at 3, 5 and 7 times the pitch. Buzzy and bright sounds look like this.
On a 1/3 octave view, pink noise shows as a flat line. That makes it the standard signal for tuning speakers and rooms.
A spike at 50Hz or 60Hz with smaller ones at 100Hz or 120Hz and up. It comes from power cables and ground loops.
A bumpy hill from about 100Hz to 4kHz with peaks called formants. They change with every vowel you say.
One very tall, narrow spike, often between 1kHz and 4kHz. In a PA system, that spike is the frequency to cut.
The analyzer uses a Fast Fourier Transform, or FFT, which looks at a short slice of sound at a time. A bigger slice separates frequencies more finely but reacts more slowly. Figures below are for a 48kHz device.
| FFT Size | Resolution | Slice Length | Good For |
|---|---|---|---|
| 2048 | 23.4 Hz | 43 ms | Fast changes, drums, quick visual feedback |
| 4096 | 11.7 Hz | 85 ms | Music and speech in general |
| 8192 | 5.9 Hz | 171 ms | The balanced default for most jobs |
| 16384 | 2.9 Hz | 341 ms | Bass notes, hum and room modes |
| 32768 | 1.5 Hz | 683 ms | Very fine pitch detail on steady sounds |
Resolution is the sample rate divided by the FFT size. At 44.1kHz each figure is about 8% finer.
From live sound to home studios to simple curiosity, seeing sound makes problems easier to find.
Play pink noise through your system, measure with the microphone on the 1/3 octave view and adjust the EQ until the bars look even.
When a PA starts to ring, the tall spike on the FFT curve shows the exact frequency to cut on your equalizer.
A spike at 50Hz or 60Hz with a ladder of harmonics points to a ground loop or a noisy power cable.
Open a song file and compare its balance with a reference track. Too much 200Hz to 400Hz sounds muddy, too little 10kHz sounds dull.
See how much bass and treble your microphone picks up, and how loud the background noise floor is with nobody talking.
Show students overtones, vowels, birdsong and the difference between noise colors, all live on a screen.
A phone or laptop microphone is not a calibrated measurement mic, but these steps get you surprisingly close.
Short answers about how the analyzer works and how to read it.
It is a tool that shows how loud each frequency in a sound is, from deep bass on the left to high treble on the right. It turns sound into a picture you can measure.
Yes. It is free with no sign-up. Microphone sound and audio files are analyzed inside your browser and never uploaded or recorded.
Browsers protect your microphone, so you have to allow access before any site can hear it. You can turn access off again in your browser settings at any time.
An FFT view shows thousands of narrow frequency points for fine detail. An RTA groups them into wider bands, here 1/3 octave, which is easier to read when balancing a room or a PA.
A-weighting turns down the very low and very high frequencies to match how human ears hear quiet sounds. It is used for noise and hearing measurements. Use Flat to see the real signal.
Noise is random, so each moment looks a little different. Turn on Average or raise Smoothing to see the steady shape underneath.
They are dBFS, meaning decibels below the loudest signal your device can capture. They are good for comparing frequencies, not for measuring loudness in dB SPL. Use the Decibel Meter for that.
Most phone and laptop microphones stop picking up sound somewhere above 15kHz. Many audio files are also cut there to save space.
Yes. Choose Audio File and open an MP3, WAV, OGG, FLAC or M4A file. It plays and shows its spectrum live, and you can mute the speakers if you only want to look.
Yes. Save PNG downloads the current display as an image, and Export CSV saves every frequency point with its current, average and peak level for a spreadsheet.
Real problems from listeners, answered and checked by the Sound Tester Audio Team.
No discussions about this tool yet. Start the first one and the team will answer it for free.