Spectrogram viewer

Make a spectrogram in three steps

1. Choose an audio or video file.
2. Pick a linear or logarithmic frequency scale and a dynamic range of 60, 80, or 100 dB.
3. Study the spectrogram with its time and frequency labels, then download it or the waveform as a PNG.

How the spectrogram is drawn

The tool mixes the channels to mono and cuts the recording into short, overlapping windows of 2,048 samples, each shaped by a Hann window so its edges fade smoothly. A Fourier transform turns every window into a spectrum, the amount of energy at each frequency, and the spectra are placed side by side as columns of color. This is called a short-time Fourier transform. At 44.1 kHz each window is about 46 milliseconds long and the frequency steps are about 21.5 Hz apart, a balance between seeing quick events and telling close notes apart. Very long files are averaged into the width of the image, so short events can blur.

Scales and dynamic range

A linear scale gives every kilohertz the same height, which is best for spotting high-frequency details such as a cutoff, hiss, or a whistle. A logarithmic scale gives every octave the same height, which matches how we hear pitch and shows the bass and melody of music more clearly. Dynamic range sets how quiet a sound can be and still appear. At 60 dB you see only the louder parts and the picture looks clean. At 100 dB faint details show up, such as the noise floor, reverb tails, and quiet hum, which is useful when you are hunting for problems.

What you can find in a spectrogram

Mains hum shows up as a steady horizontal line at 50 or 60 Hz, often with more lines at multiples above it. A sharp vertical line is a click or a pop. An MP3 made from a low-bitrate file often shows a flat ceiling around 16 kHz with nothing above it, even when the file claims 320 kbps. Some musicians have hidden pictures in the spectrum, most famously the face Aphex Twin placed in a track on the Windowlicker single. Birdsong, whale calls, and speech each have recognizable shapes, which is why spectrograms are used in teaching, ecology, and speech research.

The waveform image

The waveform shows loudness over time: tall peaks are loud moments and flat stretches are silence. It is the quickest way to see where speech starts, whether a recording hits the ceiling, or how a song builds. Download it as a PNG for a podcast cover, a video thumbnail, a social post, or a lesson slide. The spectrogram PNG keeps its frequency and time labels, so it can go straight into a report or a forum post. Both images are drawn in your browser from the decoded audio, and the file is never uploaded.