Data Lab node
Signal Spectrogram
A spectrogram of any Signal: at every update the FFT of the last Size samples becomes one row of a time × frequency matrix that keeps History seconds. Plot it as a Heatmap (rows are frequency bins labelled in Hz, the horizontal axis is time). Magnitude in the input’s unit or dB; Max Frequency crops the top.
Inputs
| Port | Type | Default |
|---|---|---|
| SignalThe Data Lab signal to analyse — a Signal node's output, a Window, or any float array with its sample rate. Multi-channel signals are handled per channel. | structure | — |
| ChannelThe channel analysed. | int | 0 |
| SizeFFT length per row in samples (power of two, ≤ 4096). | int | 256 |
| WindowWindow function per row. | int | 0 |
| OutputMagnitude in the input’s unit, or dB. | int | 0 |
| Max FrequencyCrop the rows above this frequency (Hz); 0 = up to half the sample rate. | float | 0 |
| HistoryHow many seconds of rows the matrix keeps. | float | 10 |
| Update RateRows per second: every update adds one row — this is the time resolution of the picture. | float | 20 |
Outputs
| Port | Type |
|---|---|
| Spectrogram | structure |
| Latest Bins | float |
| Peak Frequency | float |
How It Works
Signal Spectrogram builds a time × frequency matrix: at every update it takes the FFT of the last Size samples and writes it as one row, keeping History seconds of rows.
Plot draws the result as a Heatmap — rows are frequency bins labelled in Hz, the horizontal axis is time — and Plot 3D's Surface turns the same matrix into a lit heightfield. Output is magnitude in the input's unit or dB, and Max Frequency crops the top of the range so the interesting part fills the picture.
Use Cases
- See a sound — The classic audio spectrogram, for any Signal rather than only the audio engine.
- Watch a sweep — A frequency change over time is obvious here and invisible in a single spectrum.
- EEG time-frequency — How the bands rise and fall over a minute.
- Data sculpture — The matrix into Plot 3D's Surface makes a landscape you can orbit.
- Ridgeline material — A spectrogram is a matrix, and Plot's Ridgeline kind stacks matrices into the pulsar plot.
Tips
- Size trades time resolution against frequency resolution — the uncertainty is fundamental, so pick which one you need.
- dB output is almost always the more readable picture; magnitude hides everything below the loudest thing.
- Max Frequency is the fastest improvement to most spectrograms: crop away the empty top half.
- History times Update Rate is the number of rows; very long histories at high rates cost memory.
- Use a perceptual colour ramp (via Plot's Ramp) — a rainbow invents structure that is not in the data.
Open Signal Spectrogram in the interactive catalog →
Signal Spectrogram is one of 381 nodes in Grafit Pro.
Free download. Every feature for 7 days, no card. Then €99.99 once — or keep playing your projects as a Player.