Recording Audio on Linux with PulseAudio

Recording audio on Linux doesn’t require Audacity, OBS, or any heavyweight GUI. PulseAudio (and its PipeWire-compatible replacement) ships with everything you need: pactl to inspect sources and parec to capture them.

List available audio sources

pactl list sources short

Typical output on a laptop:

74   alsa_output.pci-0000_00_1f.3.analog-stereo.monitor   PipeWire   s32le 2ch 48000Hz   SUSPENDED
75   alsa_input.pci-0000_00_1f.3.analog-stereo             PipeWire   s32le 2ch 48000Hz   SUSPENDED

Field	Index	Example	Meaning
Source index	0	74	Numeric ID used with parec --device
Name	1	alsa_output...monitor	If name ends in .monitor → system output; otherwise → microphone

The .monitor suffix is the key distinction: monitor sources capture whatever is playing through your speakers, while plain sources are your microphone or line-in.

Record with parec

parec --device <source-index> --file-format=wav <output-path>

Argument	Value	Notes
--device	74, 75, etc.	The source index from pactl list sources short
--file-format	wav	Always WAV; parec only supports WAV and raw PCM
<output-path>	~/Recordings/recording_20260517_120530_<name>.wav	File written directly to disk

Record system output and mic simultaneously

Run two parec processes in parallel:

# Process 1: system output (monitor source)
parec --device 74 --file-format=wav \
  ~/Recordings/recording_20260517_120530_system.wav &

# Process 2: microphone
parec --device 75 --file-format=wav \
  ~/Recordings/recording_20260517_120530_mic.wav &

To stop both, send SIGTERM to the processes (e.g., kill %1 %2).

Key behaviors

• WAV files are written continuously during recording — no buffered finalization step.
• Files are immediately usable after the process ends. No post-processing, transcoding, or header repair is needed.
• parec records in the source’s native format (sample rate, channel count, bit depth) and writes it faithfully to the WAV container.