Spectral Physics: The Real Differences Between White, Pink, and Brown Noise

In sound synthesis and digital signal processing, random signals are categorized using color analogies derived from visible light spectra. White noise contains equal power across all frequency bands in the audible range (20 Hz to 20,000 Hz), producing a flat Power Spectral Density (PSD). Because human hearing perceives pitch logarithmically, white noise sounds biased toward upper trebles, resembling analog TV static.

Pink noise, mathematically defined as 1/f noise, reduces its power density by approximately 3 decibels per octave as frequency doubles. This -3 dB/octave attenuation matches how human critical bands process acoustic loudness, making pink noise perceive as balanced, natural, and akin to rainfall or waterfalls.

Brown noise—named after Robert Brown's description of Brownian motion, not the pigment—attenuates at 6 decibels per octave (1/f²). Because high frequencies are steeply subdued, the signal is dominated by low-frequency, velvety rumble. Many listeners find brownian spectra gentler over extended sessions because it does not stimulate the sensitive 2 kHz–5 kHz ear canal resonance where auditory fatigue frequently occurs.