Audio, ASIO, and related

The internal mix is 48 kHz stereo. Each Mixing Unit’s output device is on the Mixing Unit dialog. Headphone’s device is under Headphone in Settings.

Monitor

The Audio bar shows Post for each Input after fader and mute. Double-click a strip or use the gear to open that Input’s audio window. The window shows Pre and Post; the fader lives there.

Audio Input

Microphone and output-device loopback are regular Inputs with kind Audio. There is no separate Audio Input collection. Audio Inputs route to Mixing Units and do not participate in Scene Audio Follow. They are hidden from Scene layers and other video pickers.

Windows microphone, output-device loopback, and WASAPI playback use cpal’s WASAPI host in shared mode. Application Audio uses rsac process loopback (f32 / 48 kHz / stereo, no PCM autoconvert). An empty device id follows the current default endpoint. Output is WASAPI shared or ASIO. Application Audio lists only processes that currently have a visible window. Process capture does not fall back to endpoint loopback. ASIO input shares the driver instance with output buses. L and R each pick any input channel. macOS Core Audio playback and microphone / loopback capture use cpal. Core Audio process loopback is not implemented and fails explicitly.

OMT send

OMT Program audio leaves through AudioIngress, not the video encode worker. A slow VMX encode must not stall the 10 ms PCM cadence.