Audio Spectrogram
See an audio file's frequencies over time and save the spectrogram as PNG.
About this tool
Open an audio file (MP3, WAV, M4A, OGG, FLAC or a video's soundtrack) and its spectrogram appears, with time across the bottom, frequency up the side and loudness as colour. Choose the FFT size, a log or linear frequency axis and the quietest level shown, then save the image as a PNG.
Reading a spectrogram
A waveform shows how loud a sound is over time. A spectrogram also shows which frequencies make it up. Dark areas are quiet, and bright orange and yellow areas are loud. A pure tone draws a single horizontal line, speech shows stacked bands (the harmonics of the voice) that move as vowels change, drums appear as vertical streaks, and hiss shows up as a haze across the top. That makes spectrograms useful in many ways. Audio engineers spot mains hum at 50 or 60 Hz, clicks, clipping and noise. Music fans check whether an "HD" file is really a re-encoded MP3: a hard ceiling at about 16 kHz usually means a 128 kbps source. Birders and naturalists identify calls, and students see harmonics and formants for themselves. Under the hood, the audio is mixed to mono, cut into short overlapping slices shaped by a Hann window, and each slice runs through a fast Fourier transform. A bigger FFT size separates nearby pitches better but blurs quick events, and a smaller one does the reverse. The log axis gives each octave equal space, like a piano keyboard, while linear shows the high frequencies in more detail. Honest limits: long files are summarised into about 1,100 slices, so trim first to study short events, and the colour scale is relative, not a calibrated sound-level measurement. The audio waveform, the live spectrum analyzer and the tone generator are related, and Sample rate and bit depth explained: what 44.1 kHz and 16-bit actually mean covers digital audio basics.
Frequently asked questions
What is a spectrogram?
A picture of sound with time across, frequency up and loudness as colour. It shows which pitches are present and when.
How can I tell if an MP3 is low quality?
Look at the top of the spectrogram. A sharp cut-off around 16 kHz usually means a low-bitrate encode such as 128 kbps.
What FFT size should I use?
2048 is a good balance. Use 4096 to separate close pitches and 1024 to see fast events more sharply.
Is my audio uploaded?
No. The analysis runs in your browser and the file never leaves your device.