Skip to main content
Edit Videos

Free · private · in your browser

Create an audio spectrogram

Choose a frequency view, inspect the sound, then download the image.

How it works
  1. ImportChoose source
  2. EditSet the change
  3. PreviewReview result
  4. ExportCreate file
InputBrowser-decodable video or audioOutputPNG image

Where a waveform shows how loud a sound is, a spectrogram shows what is in it: time runs across, frequency runs up, and brightness shows how much energy sits at each frequency at each moment. It is the tool for diagnosing hum, hiss and bandwidth limits, and for anyone studying speech, music or field recordings.

Reading the picture

A steady horizontal line low in the frame is mains hum at 50 or 60 Hz, plus its harmonics stacked above it. A bright haze across the top is hiss or fan noise. Speech appears as stacked horizontal bands — the harmonics of the voice — that slide up and down with intonation, with vertical smears at consonants. A hard horizontal edge with nothing above it, usually somewhere between 15 and 16 kHz, is the fingerprint of lossy compression: that file has been through MP3 or AAC, even if it is now sitting in a WAV.

Frequency scale and window size

A linear frequency scale gives equal space to every hertz, which crams all of speech into the bottom of the image and devotes the top half to content you can barely hear. A logarithmic scale matches how hearing works and is the better default for voice and music. There is also an unavoidable trade-off between time and frequency detail: a narrow analysis window pins down exactly when something happened but smears which frequency it was, and a wide window does the reverse. Switch views rather than expecting one image to show both.

Questions, answered

How is a spectrogram different from a waveform?

A waveform shows loudness over time only. A spectrogram breaks each moment into its frequency content, so you can see hum, hiss, and which parts of the spectrum carry the sound.

How do I spot background hum?

Look for a constant horizontal line near the bottom at 50 or 60 Hz, often with fainter lines at multiples above it. A narrow filter at that frequency removes it with minimal effect on the rest.

What is the sharp cutoff near the top of my spectrogram?

That is a lossy codec’s bandwidth limit, typically between 15 and 16 kHz. It means the audio was encoded as MP3 or AAC at some point, regardless of the file’s current format.

Should I use a linear or logarithmic frequency scale?

Logarithmic for speech and music, because it matches how we hear and gives the low end the space it deserves. Linear is useful when you are hunting a specific high-frequency tone.

Can I make a spectrogram from a video?

Yes — the video’s soundtrack is analysed directly, so there is no need to extract the audio first.