Audio spectrum analyzer
A real-time FFT analyzer for your microphone input. Find hum, check a speaker's range, identify a whistle's pitch, or watch the harmonics in your voice.
How a spectrum analyzer sees a sound
Almost every real sound is a stack of simple sine waves at different frequencies. A spectrum analyzer runs a Fast Fourier Transform (FFT) that pulls the stack apart and draws one bar per frequency. Switch the partials of a 220 Hz note on and off: the left screen shows the wave shape you would see on an oscilloscope, the right shows what the analyzer draws. Turn on 1, 3 and 5 and you are halfway to a square wave; add all of them for something closer to a sawtooth.
The ear hears the lowest partial as the pitch and the rest as tone colour, or timbre. That is why a violin and a flute playing the same A sound different, and why a spectrum shows several peaks for a single note.
Practical uses
- Find mains hum: a steady spike at 50 Hz (Europe, most of Asia) or 60 Hz (North America), often with peaks at 100/150 or 120/180 Hz, points to a ground loop.
- Check speaker range: play the frequency sweep on one device and watch the peak travel across the analyzer on another. Where it fades is where the speaker runs out.
- Pitch a whistle or a drone: the readout shows the strongest frequency and the nearest musical note.
- Compare noise: play white and pink noise from the noise generator; white rises towards the right, pink stays roughly level on this log scale.
Your phone or laptop microphone has its own frequency response and usually rolls off below about 100 Hz, so treat bass readings as relative, not absolute.
Questions people ask
What does a spectrum analyzer show?
It shows how much energy a sound has at each frequency, from deep bass on the left to treble on the right. A single whistle makes one sharp peak; a voice makes a series of peaks (the harmonics); noise makes a broad hump.
Why are the numbers in dBFS and not dB SPL?
dBFS is relative to the loudest signal your device can record. Converting to real sound pressure (dB SPL) requires a calibrated measurement microphone, which phones and laptops are not.
Can I use it to find a room resonance or hum?
Yes. Play a bass sweep from the subwoofer test or pink noise and watch for peaks that stick out; a steady spike at 50 or 60 Hz, or their multiples, usually means electrical hum.
What does FFT size change?
A larger FFT splits the spectrum into narrower frequency slices, so you can separate close tones, but it reacts more slowly. 8192 gives about 5.9 Hz per slice at 48 kHz; 2048 gives about 23 Hz.
Keep testing
- Mic testLive level meter, noise floor and a 5-second record-and-playback check.
- Frequency sweepLogarithmic or linear sweeps over any range and duration.
- Noise generatorWhite, pink and brown noise with a fade-out timer.