System audio input source. Stream from a user-selected mic/line-in/interface via getUserMedia; L+R outputs are fanned out from mono sources or split from stereo. Pick the device and switch it on from the module's own faceplate — on the rack tile or in the expanded view; the permission is asked for on that click, never on patch load, and a prior grant re-opens the input on the next visit. ⚠ The last sentence here used to read "Card owns the permission prompt + device dropdown + devicechange refresh" and every clause of it is now false: the STREAM belongs to the node ($lib/ui/modules/node-audio-input-registry, #1590) and the ROSTER + the devicechange watch belong to the app ($lib/audio/input-device.svelte), so no surface unmount can stop a capture.
Brings system audio INTO the patch: it streams from a microphone, line-in, or USB interface you pick on the module and exposes the signal as a stereo L/R pair you can patch into the rack. Mental model: a live external source — sing, plug in a guitar/synth, or capture another app — and treat it like any oscillator output, sending it through filters, effects, and out to AUDIO OUT. Pick the device and switch it on from the module itself, on the rack tile or in the expanded view; permission is requested only when you enable it, not on patch load, so loading a patch never pops a mic prompt. Once granted, opening the same rack again re-opens the input on its own. The input belongs to the module, not to the panel you opened it from, so collapsing or closing that panel never stops the capture — only stopping it yourself, or deleting the module, does. Stereo handling is automatic: a stereo device feeds L and R separately, a mono source is duplicated to both sides. (Browser capture caps at a stereo pair — more than two channels per device is native-only — so only L/R are exposed.)
the faceplate
0 inputs · 2 outputs · 1 params
outputs
id
cable
what it does
audio_l_out
audio
Left channel of the selected input device (channel 1). For a mono source this carries the single channel, duplicated to the right output as well.
audio signal
audio_r_out
audio
Right channel of the selected input device (channel 2). For a mono source it carries a copy of the left channel so both sides have signal.
audio signal
params
id
label
range
default
curve
gain
Gain
0..2gain
1
linear
controls
control
what it does
Gain
Post-source level trim applied equally to both channels, 0 (silence) to 2 (×2, +6 dB), default 1 (unity). Turn it down for a hot line-in that's clipping, up for a quiet condenser mic; there is no per-channel trim.