Awesome Linux Game Tools is a series of reviews showcasing the finest tools for Linux gamers.
NoiseTorch-ng (let’s call it NoiseTorch) is billed as an “easy to use open source application for Linux with PulseAudio or PipeWire”. The software creates a virtual microphone that suppresses noise.
Noise suppression is a pretty old topic in speech processing, dating back to at least the 1970s. As the name implies, the idea is to take a noisy signal and remove as much noise as possible while causing minimum distortion to the speech of interest.
NoiseTorch uses RNNoise, a noise suppression library based on a recurrent neural network (RNN). A RNN is a class of artificial neural networks where connections between nodes can create a cycle, allowing output from some nodes to affect subsequent input to the same nodes. RNNs are particularly effective for background noise removal because they can learn patterns across time which is essential for understanding audio.
NoiseTorch isn’t only useful for gamers who are live streaming and recording. It can be used to suppress noise in a wide range of applications.
The project doesn’t provide packages for any Linux distro. Instead, it provides a compressed binary tarball so you’ll be able to get up-and-running with the minimum of fuss whatever distro you’re running. It’s a small, statically linked, self-contained binary. The binary is only for the x86 architecture. So if you want to use NoiseTorch on say an arm64 system (such as a Raspberry Pi), you’ll need to compile the source code.
There are binary and source packages in the Arch User Repository (AUR) for Arch and Arch-based distros.