Problems that are obvious in a spectrogram
A waveform shows how loud audio is over time. A spectrogram shows which frequencies are present: time runs left to right, pitch runs bottom to top, and brightness is energy. A lot of problems that are hard to describe by ear are easy to point at in a spectrogram.
- Mains hum: a thin, steady line at 50 or 60 Hz and its multiples.
- Clicks and pops: narrow vertical streaks that reach across every frequency.
- Lossy encoding: a flat ceiling where content stops dead, often between 16 and 20 kHz. Check it with the fake FLAC checker.
- Harsh sibilance: bright bursts around 5–10 kHz on every "s".
- Rumble: energy below 30 Hz you can't hear but your limiter can.
How to use it
- Add a fileDrop audio onto the page. The panel opens straight into spectrogram view.
- Zoom inScroll over the view to zoom in time. The visible range is recomputed at full detail.
- ComparePress S to flip between the spectrogram and the loudness-shaded waveform.
How it's computed
Each column is a 2048-point FFT with a Hann window, run in a background thread so the page stays responsive. Rows are spaced logarithmically from 20 Hz to the file's Nyquist frequency (half its sample rate). Colour runs from black at −110 dB through green to near-white at −10 dB relative to full scale.
Spectrum analyzer questions
Why a logarithmic frequency axis?
We hear pitch logarithmically: each octave doubles the frequency. On a log axis, bass, mids and treble get fair space instead of the top half of the picture being 11 to 22 kHz.
What does a flat line across the top mean?
A sharp edge where content stops, often around 16 to 20 kHz, usually means the file was encoded as MP3 or AAC at some point. The fake FLAC checker measures it for you.
Can I see a live microphone?
Not yet. The analyzer works on files, so you get the whole recording at once and can zoom anywhere in it.
Is my audio uploaded?
No. The FFT runs in a Web Worker on your device.