Snakey: Headphones wearing Python

PyAudio

PyAudio provides Python bindings for PortAudio v19, the cross-platform audio I/O library. With PyAudio, you can easily use Python to play and record audio on a variety of platforms, such as GNU/Linux, Microsoft Windows, and Apple macOS.

PyAudio is distributed under the MIT License.

This library was originally inspired by:

What's new

November 23, 2024

PyAudio 0.2.14 pre-compiled wheels for Microsoft Windows (32-bit and 64-bit) are now available for Python 3.13 on PyPI. Install using pip, as described below.

November 6, 2023

PyAudio 0.2.14 is a new release that fixes a build issue. Thanks to Bhanu Victor DiCara for the help.

Also, PyAudio 0.2.14 pre-compiled wheels for Microsoft Windows (32-bit and 64-bit) are now available for Python 3.12 on PyPI. Install using pip, as described below.

December 26, 2022

PyAudio 0.2.13 is a new release focused on refactoring, maintenance, and cleanup. See the CHANGELOG in the source distribution for details. Notable changes include:

October 25, 2022

PyAudio 0.2.12 pre-compiled wheels for Microsoft Windows (32-bit and 64-bit) are now available for Python 3.11 on PyPI. Install using pip, as described below.

July 18, 2022

PyAudio 0.2.12 is a new release with many updates. See the CHANGELOG in the source distribution for details. Notably, the PyAudio build process for Microsoft Windows is streamlined, using the native toolchain. Special thanks to:

Older news...

Installation

The current version is PyAudio v0.2.14 and supports Python version 3.8+. Install PyAudio using pip on most platforms. For PyAudio versions prior to v0.2.9, PyAudio distributed installation binaries, which are archived here.

Microsoft Windows

Install using pip:

python -m pip install pyaudio

Notes:


Apple macOS

Use Homebrew to install the prerequisite portaudio library, then install PyAudio using pip:

brew install portaudio
pip install pyaudio

Notes:


GNU/Linux

Use the package manager to install PyAudio. For example, on Debian-based systems:

sudo apt-get install python3-pyaudio

If the latest version of PyAudio is not available, install it using pip:

pip install pyaudio

Notes:


PyAudio Source

Source is available for download at the Python Package Index (PyPI).

Or clone the git repository:

git clone https://people.csail.mit.edu/hubert/git/pyaudio.git

To build PyAudio from source, you will also need to build PortAudio v19. See the INSTALLATION file in the PyAudio source bundle for build instructions.

Documentation & Examples

Browse the PyAudio API documentation. PyAudio roughly mirrors the PortAudio v19 API 2.0.

The PyAudio source distribution contains a set of demos. Here's a selection from that set:

"""PyAudio Example: Play a wave file."""

import wave
import sys

import pyaudio

CHUNK = 1024

if len(sys.argv) < 2:
    print(f'Plays a wave file. Usage: {sys.argv[0]} filename.wav')
    sys.exit(-1)

with wave.open(sys.argv[1], 'rb') as wf:
    # Instantiate PyAudio and initialize PortAudio system resources (1)
    p = pyaudio.PyAudio()

    # Open stream (2)
    stream = p.open(format=p.get_format_from_width(wf.getsampwidth()),
                    channels=wf.getnchannels(),
                    rate=wf.getframerate(),
                    output=True)

    # Play samples from the wave file (3)
    while len(data := wf.readframes(CHUNK)):  # Requires Python 3.8+ for :=
        stream.write(data)

    # Close stream (4)
    stream.close()

    # Release PortAudio system resources (5)
    p.terminate()
"""PyAudio Example: Record a few seconds of audio and save to a wave file."""

import wave
import sys

import pyaudio

CHUNK = 1024
FORMAT = pyaudio.paInt16
CHANNELS = 1 if sys.platform == 'darwin' else 2
RATE = 44100
RECORD_SECONDS = 5

with wave.open('output.wav', 'wb') as wf:
    p = pyaudio.PyAudio()
    wf.setnchannels(CHANNELS)
    wf.setsampwidth(p.get_sample_size(FORMAT))
    wf.setframerate(RATE)

    stream = p.open(format=FORMAT, channels=CHANNELS, rate=RATE, input=True)

    print('Recording...')
    for _ in range(0, RATE // CHUNK * RECORD_SECONDS):
        wf.writeframes(stream.read(CHUNK))
    print('Done')

    stream.close()
    p.terminate()
"""PyAudio Example: full-duplex wire between input and output."""

import sys

import pyaudio

RECORD_SECONDS = 5
CHUNK = 1024
RATE = 44100

p = pyaudio.PyAudio()
stream = p.open(format=p.get_format_from_width(2),
                channels=1 if sys.platform == 'darwin' else 2,
                rate=RATE,
                input=True,
                output=True,
                frames_per_buffer=CHUNK)

print('* recording')
for i in range(0, int(RATE / CHUNK * RECORD_SECONDS)):
    stream.write(stream.read(CHUNK))
print('* done')

stream.close()
p.terminate()
"""PyAudio Example: Play a wave file (callback version)."""

import wave
import time
import sys

import pyaudio

if len(sys.argv) < 2:
    print(f'Plays a wave file. Usage: {sys.argv[0]} filename.wav')
    sys.exit(-1)

with wave.open(sys.argv[1], 'rb') as wf:
    # Define callback for playback (1)
    def callback(in_data, frame_count, time_info, status):
        data = wf.readframes(frame_count)
        # If len(data) is less than requested frame_count, PyAudio automatically
        # assumes the stream is finished, and the stream stops.
        return (data, pyaudio.paContinue)

    # Instantiate PyAudio and initialize PortAudio system resources (2)
    p = pyaudio.PyAudio()

    # Open stream using callback (3)
    stream = p.open(format=p.get_format_from_width(wf.getsampwidth()),
                    channels=wf.getnchannels(),
                    rate=wf.getframerate(),
                    output=True,
                    stream_callback=callback)

    # Wait for stream to finish (4)
    while stream.is_active():
        time.sleep(0.1)

    # Close the stream (5)
    stream.close()

    # Release PortAudio system resources (6)
    p.terminate()
"""PyAudio Example: Audio wire between input and output. Callback version."""

import time
import sys

import pyaudio

DURATION = 5  # seconds

def callback(in_data, frame_count, time_info, status):
    return (in_data, pyaudio.paContinue)

p = pyaudio.PyAudio()
stream = p.open(format=p.get_format_from_width(2),
                channels=1 if sys.platform == 'darwin' else 2,
                rate=44100,
                input=True,
                output=True,
                stream_callback=callback)

start = time.time()
while stream.is_active() and (time.time() - start) < DURATION:
    time.sleep(0.1)

stream.close()
p.terminate()
"""PyAudio Example: macOS-only: Play a wave file with channel maps."""

import wave
import sys

import pyaudio


CHUNK = 1024

if len(sys.argv) < 2:
    print(f'Plays a wave file. Usage: {sys.argv[0]} filename.wav')
    sys.exit(-1)

# standard L-R stereo
# channel_map = (0, 1)

# reverse: R-L stereo
# channel_map = (1, 0)

# no audio
# channel_map = (-1, -1)

# left channel audio --> left speaker; no right channel
# channel_map = (0, -1)

# right channel audio --> right speaker; no left channel
# channel_map = (-1, 1)

# left channel audio --> right speaker
# channel_map = (-1, 0)

# right channel audio --> left speaker
channel_map = (1, -1)
# etc...
try:
    stream_info = pyaudio.PaMacCoreStreamInfo(
        flags=pyaudio.PaMacCoreStreamInfo.paMacCorePlayNice,
        channel_map=channel_map)
except AttributeError:
    print(
        'Could not find PaMacCoreStreamInfo. Ensure you are running on macOS.')
    sys.exit(-1)

print('Stream Info Flags:', stream_info.flags)
print('Stream Info Channel Map:', stream_info.channel_map)

with wave.open(sys.argv[1], 'rb') as wf:
    p = pyaudio.PyAudio()
    stream = p.open(
        format=p.get_format_from_width(wf.getsampwidth()),
        channels=wf.getnchannels(),
        rate=wf.getframerate(),
        output=True,
        output_host_api_specific_stream_info=stream_info)

    # Play stream
    while len(data := wf.readframes(CHUNK)):  # Requires Python 3.8+ for :=
        stream.write(data)

    stream.close()
    p.terminate()

License

PyAudio is distributed under the MIT License:

Copyright (c) 2006 Hubert Pham

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

Acknowledgements

Many thanks to Andrew Baldwin, Alex ThreeD, Timothée Lecomte, Frank Samuelson, Matthieu Brucher, Chris Stawarz, Barry Walker, Bob Jamison, Danilo J. S. Bellini, Bastian Bechtold, Christoph Gohlke, Sebastian Audet, Jason Roehm, Tony Jacobson, Sami Liedes, Michael Graczyk, Blaise Potard, Matthias Schaff, Timothy Port, Sean Zimmermann, Jason Hihn, Eiichi Takamori, Matěj Cepl, Vasily Zakharov, Artur Janowiec, and Bhanu Victor DiCara for their much appreciated suggestions and patches—as well as to others who have written to say hello!

Special thanks to Felipe Sateler and Justin Mazzola Paluska for Debian/Ubuntu packaging help.

The development of PyAudio was funded in part by the Cambridge-MIT Institute and T-Party.

Contact

For help with PyAudio, your best bet for a timely response is to consult or post to StackOverflow.

Comments, suggestions, and patches welcomed. Send mail to my first name at mit.edu.