Skip to content

Non-Stationary Noise

Definition

Interference whose spectral content changes over time: a vehicle passing, a crowd noise that rises and falls, or a door slamming. Adaptive filtering is required because a fixed noise estimate based on a single silent interval will not model the interference throughout the recording.

Defining trait
Spectral content changes over time
Typical examples
Passing vehicle, rising and falling crowd noise, a door slam
Mitigation approach
Adaptive filtering
Contrast
Stationary noise, e.g. a constant hum, keeps a fixed spectral profile

Common questions

Why can't a single silent segment be used to model this kind of noise?+

A noise profile taken from one quiet moment only captures the interference present at that instant, so applying it across the whole recording either under-suppresses noise that changes character later or removes speech energy where the noise profile no longer matches.

What techniques handle non-stationary noise in forensic audio enhancement?+

Adaptive and spectral-subtraction filters that continuously re-estimate the noise floor, along with Wiener filtering and, increasingly, neural denoising models trained to track changing background interference, are used to keep pace with the shifting noise.

Related terms

PESQ / STOI
Perceptual Evaluation of Speech Quality and Short-Time Objective Intelligibility: instrumental measures that predict listener intelligibility without needing human subjects. STOI scores between...
Spectral Subtraction
An enhancement method that estimates the noise spectrum during a speech-free interval and subtracts it from every subsequent frame. Simple and effective...
Stationary Noise
Noise whose spectral characteristics remain approximately constant over the duration of a recording. Tape hiss, HVAC fan noise, and mains hum are...
SWGDE
Scientific Working Group for Digital Evidence. A US multi-agency body that publishes consensus best-practice documents for digital forensic disciplines, including image authentication,...
Wiener Filter
A minimum mean-squared-error filter that computes a frequency-dependent gain function from the signal-to-noise ratio in each frequency band. Less aggressive than spectral...

Explained in

Your journey to becoming a forensic professional starts here.

Practice with mock tests, learn from structured notes, and get your questions answered by a global forensic community, all in one place.