Fact · Speech recognition
The python-sounddevice package records audio from the sound device into a NumPy array with its rec() function.
Checked against its source on October 17, 2026
In the source's own words
“To record audio data from your sound device into a NumPy array, you can use rec()”
Play and Record Sound with Python: Usage (python-sounddevice), Matthias Geier, python-sounddevice documentation, n.d., Recording.
Open the original- What it measures
- The usage page gives rec() a duration and a samplerate, and you pass the sampling rate yourself. The package sits on the PortAudio library.
- What it does not show
- Recording needs the sample rate your model expects. Set the samplerate argument explicitly rather than relying on the device default.