Audio configuration

Open More, Settings, Voice from the voice dock. This page reads and changes the audio engine’s settings. If the engine is unavailable, the page reports that instead of offering controls that cannot work.

Devices and volume

Choose an Input device for your microphone and an Output device for speakers or headphones. System default follows the operating system’s choice. Microphone and speaker volume sliders are separate.

On supported Windows builds, Exclusive microphone mode takes the device directly from its driver and can prevent another application from using it at the same time.

Activation and voice gate

Choose Voice activation, Continuous, or Push to talk. Set the push-to-talk shortcut when that mode is selected.

Auto calibrate asks you to speak naturally for about five seconds and tunes the gate. Manual calibrate exposes the Open and Close markers on the level meter. The gate’s close threshold and hold prevent rapid switching between transmitting and silence.

Use Hear yourself to record a short sample through the current filters and listen back before a call.

Processing

Auto gain adjusts microphone level for consistent speech. Noise suppression offers the algorithms included in your build, including RNNoise, DeepFilterNet, OMLSA + IMCRA, and spectral subtraction where available. Choose a setting that keeps speech clear on your hardware; compare it with noise suppression off using the sample recorder.

Transmission

Higher Quality uses more bandwidth. Audio per packet trades packet overhead against added latency. Force TCP audio sends voice through the control tunnel when UDP is blocked; use the normal UDP path when it works.

Additional controls appear for advanced user modes. Choose your mode in Settings, Advanced before looking for expert engine and network settings.

If voice fails entirely, start with Audio troubleshooting.