Noise suppression removes the steady and intermittent background sounds around a talker — fans, traffic, keyboard taps, a barking dog — while leaving speech intact and natural. Classic approaches estimate the noise spectrum during pauses and subtract it, which works for stationary hums but smears speech and struggles with sudden sounds. Modern systems are deep neural networks (RNNoise, Krisp, NVIDIA RTX Voice) trained on huge mixes of speech and noise; they separate voice from everything else far more aggressively, even pulling a clear talker out of a noisy cafe. The tradeoffs are CPU or GPU cost and the risk of over-suppression producing a processed, underwater quality, so tuning matters.
The book · Volume 4 of 7
Audio for Video: A Complete Guide to Sound in Video Products: From PCM and Codecs to Loudness, Accessibility, and Real-Time Voice Processing
Noise suppression is the stage users notice first when it fails: muffled speech, pumping, or a fan that never goes away. Volume 4 covers noise suppression in detail as part of real-time voice processing, together with AEC, AGC, VAD and the Opus codec that follows them in the chain. Written by Nikolay Sapunov, CEO at Fora Soft.
Users complaining about keyboard and fan noise on your calls, and noise suppression isn't catching it?
Send us a few problem recordings and we will recommend a noise suppression setup, classical, RNNoise or a commercial SDK, with the CPU budget per call. 250+ video products shipped since 2005.
