diff --git a/README.rst b/README.rst
index 6207d27..b4db676 100644
--- a/README.rst
+++ b/README.rst
@@ -85,11 +85,15 @@ For validation, run::
     optional arguments:
       -h, --help     show this help message and exit
 
+`python-sounddevice <https://python-sounddevice.readthedocs.io>` is used by default as the audio library.
+This should work out-of-the box on Linux, Windows and macOS, however, other choices are available.
+
 On, Windows you may download the `portaudio` library from `MinGW <https://packages.msys2.org/base/mingw-w64-portaudio>`_.
 Then, you should specify the DLL using the following command-line flag::
 
     -l AUDIO_LIBRARY, --audio-library AUDIO_LIBRARY
-                        File name of PortAudio shared library.
+
+The AUDIO_LIBRARY can be 'alsa', 'sd' (the default) or the file name for the PortAudio shared library.
 
 
 Calibration
diff --git a/amodem/__main__.py b/amodem/__main__.py
index 2243c4b..62218c4 100644
--- a/amodem/__main__.py
+++ b/amodem/__main__.py
@@ -169,8 +169,9 @@ def create_parser(description, interface_factory):
     for sub in subparsers.choices.values():
         sub.add_argument('-c', '--calibrate', nargs='?', default=False,
                          metavar='SYSTEM', help=calibration_help)
-        sub.add_argument('-l', '--audio-library', default='libportaudio.so',
-                         help='File name of PortAudio shared library.')
+        sub.add_argument('-l', '--audio-library', default='sd',
+                         help="'alsa', 'sd' or file name of "
+                         "PortAudio shared library")
         sub.add_argument('-z', '--zlib', default=False, action='store_true',
                          help='Use zlib to compress/decompress data.')
         g = sub.add_mutually_exclusive_group()
@@ -234,9 +235,12 @@ def interface_factory():
         import pylab  # pylint: disable=import-error,import-outside-toplevel
         args.pylab = pylab
 
-    if args.audio_library == 'ALSA':
+    if args.audio_library in ['alsa', 'ALSA']:
         from . import alsa  # pylint: disable=import-outside-toplevel
         interface = alsa.Interface(config)
+    elif args.audio_library == 'sd':
+        from . import sd  # pylint: disable=import-outside-toplevel
+        interface = sd.Interface(config)
     elif args.audio_library == '-':
         interface = _Dummy()  # manually disable PortAudio
     elif args.command == 'send' and args.output is not None:
diff --git a/amodem/sd.py b/amodem/sd.py
new file mode 100644
index 0000000..490039f
--- /dev/null
+++ b/amodem/sd.py
@@ -0,0 +1,69 @@
+"""
+Code which adds sounddevice lib support for interfaces, recording and playing.
+"""
+
+import logging
+import sounddevice as sd
+
+log = logging.getLogger(__name__)
+
+
+class Interface:
+
+    def __init__(self, config):
+        self.config = config
+        bits_per_sample = config.bits_per_sample
+        assert bits_per_sample == 16
+
+    def __enter__(self):
+        return self
+
+    def __exit__(self, *args):
+        pass
+
+    def recorder(self):
+        return Recorder(self)
+
+    def player(self):
+        return Player(self)
+
+
+class Recorder:
+    def __init__(self, lib):
+        self.bufsize = 4096
+
+        self.audio_stream = sd.RawInputStream(
+            samplerate=lib.config.Fs,
+            blocksize=self.bufsize,
+            channels=1, dtype='int16')
+        self.audio_stream.start()
+
+    def read(self, size):
+        data, overflowed = self.audio_stream.read(size)
+        if overflowed:
+            raise OverflowError("Overflow reading from audio device")
+        return data
+
+    def close(self):
+        self.audio_stream.stop()
+        self.audio_stream.close()
+
+
+class Player:
+    def __init__(self, lib):
+        self.buffer_length_ms = 10
+        self.buffer_size = int(lib.config.Fs * (self.buffer_length_ms / 1000))
+
+        self.audio_stream = sd.RawOutputStream(
+            samplerate=lib.config.Fs,
+            blocksize=self.buffer_size,
+            channels=1, dtype='int16')
+
+        self.audio_stream.start()
+
+    def write(self, data):
+        self.audio_stream.write(data)
+
+    def close(self):
+        self.audio_stream.stop()
+        self.audio_stream.close()
diff --git a/setup.py b/setup.py
index 4b9cf05..0b7400d 100644
--- a/setup.py
+++ b/setup.py
@@ -24,7 +24,7 @@ def run_tests(self):
     packages=['amodem'],
     tests_require=['pytest'],
     cmdclass={'test': PyTest},
-    install_requires=['numpy'],
+    install_requires=['numpy','sounddevice'],
     platforms=['POSIX'],
     classifiers=[
         'Development Status :: 5 - Production/Stable',