GIL-less Portaudio Streams for Python

pastream builds on top of portaudio and the excellent sounddevice python bindings to provide some more advanced functionality right out of the box. Note that in addition to the pastream library, pastream includes a command line application for playing and recording audio files.

Documentation:

http://pastream.readthedocs.io/

Source code repository and issue tracker:

http://github.com/tgarc/pastream/

Features

GIL-less Audio Callbacks

Having the portaudio callback implemented in C means audio interrupts can be serviced quickly and reliably without ever needing to acquire the Python Global Interpreter Lock (GIL). This is crucial when working with libraries like Pillow which may greedily grab and hold the GIL subsequently causing audio overruns/underruns.

Input Stream iterators

Efficiently retrieve live audio capture data through an iterable. As simple as:

import pastream as ps
for chunk in ps.chunks():
    process(chunk)

See pastream.chunks and pastream.InputStream.chunks method.

Built-in support for working with SoundFiles and numpy ndarrays

Seamless support for playback/recording of numpy ndarrays, generic buffer types, and SoundFiles.

Reader/Writer Threads

pastream simplifies the process of implementing stream reader and writer threads to manipulate and/or generate data in the background while leaving the main thread free for higher level management tasks.

External Dependencies

There are a few compiled libraries pastream requires which may need to be installed separately depending on your operating system. Windows users are luckiest, they can skip this section entirely.

libffi (Linux/Unix/MacOSX):

Under Linux/Unix/MacOSX platforms you'll need to install the ffi library. (For Windows users, ffi is already included with the python cffi package.) libffi is available through most package managers:

$ yum install libffi-devel # Red-hat/CentOS Linux
$ apt-get install libffi-dev # Ubuntu/debian derivatives
$ brew install libffi # Homebrew on OSX

More information on installing libffi is available in the cffi documentation here.

PortAudio and libsndfile (Linux/Unix):

Linux and Unix users will also need to install a recent version of the PortAudio and libsndfile libraries. (For Windows and OSX, the sounddevice and soundfile python packages include prebuilt versions for you.) You can either install the latest available from your package manager (e.g. apt-get install libportaudio2 libsndfile for debian/raspbian) or install the latest stable build from the package website (Recommended).

Installation

Once the above dependencies have been resolved, you can install pastream using pip:

$ pip install pastream

Building From Source

Clone pastream with the --recursive flag:

$ git clone --recursive http://github.com/tgarc/pastream

Or, if you already have a checkout:

$ cd <path/to/checkout>
$ git submodule update --init

Finally, do a pip install from your local working copy:

$ pip install <path/to/checkout>

Building Documentation

Documentation for pastream can be easily generated in a wide variety of formats using Sphinx. Just follow the steps below.

Checkout the repository and cd into it:

$ git clone http://github.com/tgarc/pastream
$ cd pastream

Install documentation dependencies using requirements file:

$ pip install -r docs/requirements.txt

Then use the included Makefile/make.bat to generate documentation. (Here we output to the html format):

$ cd docs
$ make html

Examples

Record one second of audio to memory, then play it back:

import pastream as ps

# Use *with* statements to auto-close the stream
with ps.DuplexStream() as stream:
    out = stream.record(int(stream.samplerate), blocking=True)
    stream.play(out, blocking=True)

Playback 10 seconds of a file, adding zero padding if the file is shorter, and record the result to memory:

import pastream as ps, soundfile as sf

with sf.SoundFile('my-file.wav') as infile, ps.DuplexStream.from_file(infile) as stream:
    out = stream.playrec(infile, frames=10 * int(stream.samplerate), pad=-1, blocking=True)

Grab (real) frequency transformed live audio stream with 50% overlap:

import pastream as ps, numpy as np

chunksize = 1024
window = np.hanning(chunksize)
for x_l in ps.chunks(chunksize, overlap=chunksize//2, channels=1):
    X_l = np.fft.rfft(x_l * window)

Generate a pure tone on-the-fly

import time
import pastream as ps
import numpy as np

# A simple tone generator
def tone_generator(stream, buffer, f, loop=False):
    fs = stream.samplerate

    # Create a time index
    t = 2*np.pi*f*np.arange(len(buffer), dtype=stream.dtype) / fs

    # Loop until the stream stops
    while not stream.finished:
        frames = buffer.write_available
        if not frames:
            time.sleep(0.010)
            continue

        # Get the write buffers directly to avoid making any extra copies
        frames, part1, part2 = buffer.get_write_buffers(frames)

        out = np.frombuffer(part1, dtype=stream.dtype)
        np.sin(t[:len(out)], out=out)

        if len(part2):
            # part2 will be nonempty whenever we wrap around the end of the ring buffer
            out = np.frombuffer(part2, dtype=stream.dtype)
            np.sin(t[:len(out)], out=out)

        # flag that we've added data to the buffer
        buffer.advance_write_index(frames)

        # advance the time index
        t += 2*np.pi*f*frames / fs

with ps.OutputStream(channels=1) as stream:
    # Set our tone generator as the source and pass along the frequency
    freq = 1000
    stream.set_source(tone_generator, args=(freq,))

    # Busy-wait to allow for keyboard interrupt
    stream.start()
    while stream.active:
        time.sleep(0.1)

See also the included examples under /examples.

Command Line Application

Once installed, the pastream application should be callable from your command line. If you're familiar with SoX you'll notice that some of the command line syntax is quite similar. Here are a few examples to help get you started.

Display the help file:

$ pastream -h

List available audio devices:

$ pastream -l

Simultaneous play and record from the default audio device:

$ pastream input.wav output.wav

Pipe input from sox using the AU format and record the playback:

$ sox -n -t au - synth sine 440 | pastream - output.wav

Play a RAW file:

$ pastream -c1 -r48k -e=pcm_16 output.raw

Record 10 minutes of audio at 48kHz:

$ pastream null output.wav -r48k -d10:00