Microphone access in Python

Question

Can I access a users microphone in Python?

Sorry I forgot not everyone is a mind reader: Windows at minimum XP but Vista support would be VERY good.

Useful answers in this more recent question: http://stackoverflow.com/questions/1936828/how-get-sound-input-from-microphone-in-python-and-process-it-on-the-fly — Craig McQueen, Dec 20 '09 at 23:46

score 18 · Answer 1 · answered Oct 29 '08 at 15:53

18

I got the job done with pyaudio

It comes with a binary installer for windows and there's even an example on how to record through the microphone and save to a wave file. Nice! I used it on Windows XP, not sure how it will do on Vista though, sorry.

answered Oct 29 '08 at 15:53

Jeffrey Martinez

4,384
2
31
36

1

"Note that PyAudio currently only supports blocking-mode audio I/O. PyAudio is still super-duper alpha quality." from website – Dustin Getz Jul 23 '09 at 18:20

score 4 · Accepted Answer · answered Oct 11 '08 at 07:47

4

Best way to go about it would be to use the ctypes library and use WinMM from that. mixerOpen will open a microphone device and you can read the data easily from there. Should be very straightforward.

answered Oct 11 '08 at 07:47

Serafina Brocious

30,433
12
89
114

3

Do you think you can provide some sample code for this? I can call mixerGetNumDevs but i'm not sure how to get from there to mixerOpen or for reading levels. TIA – Jared Apr 10 '09 at 04:51

score 2 · Answer 3 · answered Oct 14 '08 at 19:07

2

You might try SWMixer.

answered Oct 14 '08 at 19:07

Bill Barksdale

2,466
1
14
6

score 1 · Answer 4 · answered May 11 '23 at 02:04

Just as an update to Martinez' answer above, I also used pyAudio (latest edition is 0.2.13 since Dec 26 2022).

Here's how to install pyaudio on windows (I did it in a virtual environment):

pip install pyaudio   # as of python 3.10 this should download a wheel

Once you have it installed, assuming you want to record into a 16-bit wave file, this snippet adapted from the documentation should help:

import wave      # this will probably also need to be installed
import pyaudio

RATE = 16000
FORMAT = pyaudio.paInt16 # 16-bit frames, ie audio is in 2 bytes
CHANNELS = 1             # mono recording, use 2 if you want stereo
CHUNK_SIZE = 1024        # bytes
RECORD_DURATION = 10     # how long the file will be in seconds

with wave.open("recording.wav", "wb") as wavefile:
    p = pyaudio.PyAudio()
    wavefile.setnchannels(CHANNELS)
    wavefile.setsampwidth(p.get_sample_size(FORMAT))
    wavefile.setframerate(RATE)

    stream = p.open(format=FORMAT, channels=CHANNELS, rate=RATE, input=True)
    for _ in range(0, RATE // CHUNK_SIZE * RECORD_DURATION):
        wavefile.writeframes(stream.read(CHUNK_SIZE))
    stream.close()

    p.terminate()

I wasn't able to use with for the stream handler, but this might be possible in future editions of pyAudio.

Microphone access in Python

4 Answers4

Linked