Noise torch It is a real-time microphone noise suppression application for Linux, which can filter out unwanted background noise, such as mechanical keyboards, computer fans, trains, etc.It currently only supports PulseAudio, but PipeWire support It is planned to be released in a future version.
Keep simplicity in mind when building the user interface of an application. If there is only 1 microphone, you just need to launch the app and click “Load NoiseTorch”. After this is done, the application will create a virtual microphone called “NoiseTorch Microphone”:
You can select this virtual microphone in any application, such as Zoom, Mumble, Skype, Slack, Discord, TeamSpeak, Microsoft Teams, etc. to filter out background noise.
If you have multiple microphones, click “Select Microphone” and select your microphone from the list.
Related: Hushboard mutes your microphone while typing
NoiseTorch only sends voice input when it detects voice. If this does not apply to you, you can use the “Voice Activation Threshold” option to solve this problem. The developer states that “using a decent microphone, you can turn it up to 95%”, but if the microphone is interrupted during an audio conference, you need to lower this value until it works for you.
After changing the settings, you need to restart NoiseTorch to use them.
The app is also able to perform output filtering (“filter headphones” option). This is a customer-only function, but you can use it even if you are not a NoiseTorch customer. However, the developer pointed out that in this case, “legally you must feel sad” “️.
It is worth mentioning that NoiseTorch introduces some latency (approximately 10 milliseconds), which is very low, so this should not be a problem. However, output filtering introduces a delay of about 100ms, which is okay in most cases, but not for competitive games.
You can use to test without actually having to talk to others
gst-launch-1.0 pulsesrc ! pulsesink Used for live microphone output (echo test).
Another Linux GUI for removing background noise in real time is Cadmus, So if you are not satisfied with NoiseTorch, please give it a try.
You may also be interested in the following: How to enable echo/noise cancellation of microphone input on the Linux desktop (PulseAudio).Noise torch can use Combined with PulseAudio echo cancellation module. But in my case, using NoiseTorch is enough.
Using the official NoiseTorch installation instructions, copy the application binaries to
~/.local/bin.If you don’t have this directory in your PATH (e.g.
~/.profile), add it here, and logout/login. Or run the application directly from its installation directory:
You can also use systemd to automatically start NoiseTorch.see Instructions here.