Audio Interface

Introduction

The gnuradio audio component provides gr::audio::source and gr::audio::sink blocks. The audio blocks stream floating point samples to and from audio hardware.

The gr-audio component will be built automatically when gnuradio-runtime is enabled. Support for underlying audio architectures depends on OS and installed libraries. At the time of writing, gr-audio supports OSS, ALSA, Jack, Portaudio, Audiounit, and Winmm.

At runtime, gr-audio will automatically select from the available architectures. The user can override the selection via configuration file by setting "audio_module" to one of the following strings:

• oss
• alsa
• jack
• portaudio
• osx
• windows

See gr-audio.conf for an example.

Import this package with:

from gnuradio import audio

See the Doxygen documentation for details about the blocks available in this package. A quick listing of the details can be found in Python after importing by using:

help(audio)

Usage

For an audio source, a typical ArgumentParser option and it's use looks like:

parser.add_argument("-O", "--audio-output", default="",
                   help="pcm device name.  E.g., hw:0,0 or surround51 or /dev/dsp")
audio_rate = 32e3
audio_sink = audio.sink (int (audio_rate), options.audio_output)

Similarly, an audio sink would have a typical ArgumentParser option and its use would look like:

parser.add_argument("-I", "--audio-input", default="",
                  help="pcm input device name.  E.g., hw:0,0 or /dev/dsp")
audio_rate = 32e3
audio_source = audio.source(int(audio_rate), audio_input)

Adding a New Audio Machine

There may come a time when we need to define a new audio machine type besides those currently supported. To do this, we have to follow a simple pattern to add it to the list of potential machines GNU Radio can use.

1. Add a new directory in gr-audio/lib for the new machine name, like the alsa, oss, etc. that are already there.
2. Follow the pattern of the other machines to create the class structure for both a source and sink implementation for the machine.
3. Make sure to add the factory function for both the new source and sink classes. Like in the ALSA sink case, we have:

sink::sptr
alsa_sink_fcn(int sampling_rate,
              const std::string &device_name,
              bool ok_to_block)
{
  return sink::sptr
    (new alsa_sink(sampling_rate, device_name, ok_to_block));
}

1. Add the new source and sink factory functions to the audio registration list in gr-audio/lib/audio_registry.cc and audio_registry.h files. Select the appropriate registry priority, HIGH if you want this to be the default when using 'auto'. For Linux audio systems, we generally want to default to ALSA, fall back on OSS, but have other machine interfaces defined as MED priority. For example, in the .cc file for ALSA, OSS, and PortAudio:

#ifdef ALSA_FOUND
        s_registry.push_back(register_source(REG_PRIO_HIGH, "alsa", alsa_source_fcn));
#endif /* ALSA_FOUND */
 
#ifdef OSS_FOUND
        s_registry.push_back(register_source(REG_PRIO_LOW, "oss", oss_source_fcn));
#endif /* OSS_FOUND */
 
#ifdef PORTAUDIO_FOUND
        s_registry.push_back(register_source(REG_PRIO_MED, "portaudio", portaudio_source_fcn));
#endif /* PORTAUDIO_FOUND */

1. Follow the examples in the gr-audio/lib/CMakeLists.txt file for the different machine types to add the new one, including the add_definitions to provide the new YOURMACH_FOUND used in the audio_registry files.