Digital Audio & DSP 8 min read Updated: October 2026

Fast Fourier Transform (FFT) in Audio: How Spectral Visualizers Work

Every real-time spectrum analyzer, parametric EQ display, and audio visualizer relies on the Fast Fourier Transform (FFT) algorithm, developed in 1965 by J.W. Cooley and John Tukey to compute the Discrete Fourier Transform (DFT) in $O(N \log N)$ operations.

Time Domain vs Frequency Domain

An audio file stored on disk exists in the time domain: a series of amplitude values over time. The Fourier transform mathematically decomposes this complex waveform into an infinite sum of constituent sine and cosine waves, revealing the amplitude and phase of each frequency band (the frequency domain).

FFT Windowing & Frequency Resolution

Because real-time audio is continuous, the incoming stream is sliced into discrete chunks called window frames (e.g. 1024, 2048, or 4096 samples). Windowing functions (such as Hann, Blackman, or Hamming) taper the edges of the buffer to zero, preventing spectral leakage artifacts.

The frequency resolution of each analyzer bin is governed by:

$$\Delta f = \frac{f_s}{N}$$

Bin bandwidth: at 44.1 kHz with a 2048-point FFT, each frequency bin spans 21.5 Hz

← Back to Knowledge Base Index Explore Web Audio Suite →