Spectrum Analyzer

A frequency-domain visualizer that runs an FFT on the signal. It sits in the chain as a pass-through, so audio flows through unchanged.

spectrum

Spectrum Analyzer - Frequency-domain FFT display.

Param	Type	Description
in	signal	Input signal
name	string	Display label (optional)
opts	record	Options (optional)

Options:

Option	Type	Default	Description
width	number / string	200	Width in pixels, or "100%" for full width
height	number / string	50	Height in pixels, or "100%" for full height
fft	number	1024	FFT size: 256, 512, 1024, or 2048
logScale	boolean	false	Logarithmic frequency axis
minDb	number	-90	Minimum dB for display range
maxDb	number	0	Maximum dB for display range

Displays a real-time frequency spectrum. Larger FFT sizes give better frequency resolution but slower updates.

// Basic spectrum
osc("saw", 220) |> spectrum(%) |> out(%, %)

fft

FFT window size. Larger values resolve closely-spaced frequencies better but slow down the update rate. Valid sizes: 256, 512, 1024 (default), 2048.

// High resolution FFT
osc("saw", 55) |> spectrum(%, "bass detail", {fft: 2048, width: 400}) |> out(%, %)

logScale

Use a logarithmic frequency axis. Linear (default) spaces frequencies evenly; logarithmic gives more detail to low frequencies, matching how we perceive pitch.

// Log scale matches human pitch perception
osc("saw", 110) |> lp(%, 2000) |> spectrum(%, "filtered", {logScale: true}) |> out(%, %)

minDb

Lower bound of the dB display range. Values below this are clamped to the bottom of the plot. Useful for cutting visual noise floor.

maxDb

Upper bound of the dB display range. Loud signals above this are clamped to the top.

// Tight dB window for a hot signal
osc("saw", 110) |> spectrum(%, {minDb: -40, maxDb: 0}) |> out(%, %)

Related: waterfall, oscilloscope