ZapKit

Audio spectrogram

See sound the way audio engineers do: a heat map of which frequencies are present at every moment, to spot hum, hiss, whistles and the shape of voices.

What is a spectrogram?

A spectrogram is a picture of sound that shows frequency on one axis, time on the other, and loudness as colour. It reveals things you can't see in a waveform: the pitch of a voice, the notes of an instrument, a steady electrical hum, or a high-pitched whine hidden under speech.

Open a file to draw its spectrogram. Choose the highest frequency to show, a musical log scale or a linear scale, and how faint a sound should still appear. Download the image as a PNG to share or compare.

How to use the Audio spectrogram

  1. 1.Open a recording

    Audio, or a video with sound.

  2. 2.Adjust the view

    Highest frequency, log or linear scale, and how quiet a sound to show.

  3. 3.Read or download

    Save the spectrogram as a PNG.

Example: finding hum in a recording

An interview recorded near a power adapter:

Range
2 kHz, log scale
Seen
a bright steady line at 50 Hz with fainter lines at 100 and 150 Hz
Meaning
mains hum and its harmonics

Now you know to filter out 50 Hz and its multiples, or re-record away from the adapter.

Reading a spectrogram

Time runs left to right and frequency rises from bottom to top; brighter colour means louder. A steady horizontal line is a constant tone, such as 50 or 60 Hz mains hum near the bottom. A bright haze across all frequencies is broadband noise, and stacked curved lines are the harmonics of a voice or instrument.

The log scale gives more room to low frequencies, where pitch and most musical detail live; the linear scale spreads all frequencies evenly, which suits high-pitched noise.

Frequently asked questions

How do I find a hum?

Look for a bright horizontal line at 50 or 60 Hz, often with repeats at multiples above it.

What does the dB setting change?

How quiet a sound still shows. Lower values reveal faint noise; higher values show only strong sounds.

Can it detect a fake (upscaled) MP3?

Often. Low-quality MP3s show a sharp cut-off around 16 kHz with the Full range view.

Is my file uploaded?

No. The analysis runs locally.

More Audio & video tools