Stationary Noise
Definition
Noise whose spectral characteristics remain approximately constant over the duration of a recording. Tape hiss, HVAC fan noise, and mains hum are typical examples. Stationary noise is tractable by spectral subtraction and Wiener filtering.
- Property
- Spectral characteristics roughly constant over the recording
- Examples
- Tape hiss, HVAC fan noise, mains hum
- Tractable by
- Spectral subtraction, Wiener filtering
- Field
- Audio enhancement and speech intelligibility
Common questions
Why does stationary noise respond better to filtering than non-stationary noise?+
Because its spectral profile stays consistent, a filter can estimate the noise spectrum from a quiet segment of the recording and subtract that same profile throughout, whereas noise that changes character over time defeats a fixed estimate.
Can stationary noise removal introduce its own artefacts into forensic audio?+
Yes. Aggressive spectral subtraction can leave a metallic warble known as musical noise and can also remove low-level speech components that overlap the noise spectrum, so enhancement is applied conservatively and the original recording is preserved for comparison.
Related terms
- Non-Stationary Noise
- Interference whose spectral content changes over time: a vehicle passing, a crowd noise that rises and falls, or a door slamming. Adaptive...
- 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...
- 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...