1
2:mod:`ossaudiodev` --- Access to OSS-compatible audio devices
3=============================================================
4
5.. module:: ossaudiodev
6   :platform: Linux, FreeBSD
7   :synopsis: Access to OSS-compatible audio devices.
8
9
10.. versionadded:: 2.3
11
12This module allows you to access the OSS (Open Sound System) audio interface.
13OSS is available for a wide range of open-source and commercial Unices, and is
14the standard audio interface for Linux and recent versions of FreeBSD.
15
16.. Things will get more complicated for future Linux versions, since
17   ALSA is in the standard kernel as of 2.5.x.  Presumably if you
18   use ALSA, you'll have to make sure its OSS compatibility layer
19   is active to use ossaudiodev, but you're gonna need it for the vast
20   majority of Linux audio apps anyway.
21
22   Sounds like things are also complicated for other BSDs.  In response
23   to my python-dev query, Thomas Wouters said:
24
25   > Likewise, googling shows OpenBSD also uses OSS/Free -- the commercial
26   > OSS installation manual tells you to remove references to OSS/Free from the
27   > kernel :)
28
29   but Aleksander Piotrowsk actually has an OpenBSD box, and he quotes
30   from its <soundcard.h>:
31   >  * WARNING!  WARNING!
32   >  * This is an OSS (Linux) audio emulator.
33   >  * Use the Native NetBSD API for developing new code, and this
34   >  * only for compiling Linux programs.
35
36   There's also an ossaudio manpage on OpenBSD that explains things
37   further.  Presumably NetBSD and OpenBSD have a different standard
38   audio interface.  That's the great thing about standards, there are so
39   many to choose from ... ;-)
40
41   This probably all warrants a footnote or two, but I don't understand
42   things well enough right now to write it!   --GPW
43
44
45.. seealso::
46
47   `Open Sound System Programmer's Guide <http://www.opensound.com/pguide/oss.pdf>`_
48      the official documentation for the OSS C API
49
50   The module defines a large number of constants supplied by the OSS device
51   driver; see ``<sys/soundcard.h>`` on either Linux or FreeBSD for a listing.
52
53:mod:`ossaudiodev` defines the following variables and functions:
54
55
56.. exception:: OSSAudioError
57
58   This exception is raised on certain errors.  The argument is a string describing
59   what went wrong.
60
61   (If :mod:`ossaudiodev` receives an error from a system call such as
62   :c:func:`open`, :c:func:`write`, or :c:func:`ioctl`, it raises :exc:`IOError`.
63   Errors detected directly by :mod:`ossaudiodev` result in :exc:`OSSAudioError`.)
64
65   (For backwards compatibility, the exception class is also available as
66   ``ossaudiodev.error``.)
67
68
69.. function:: open(mode)
70              open(device, mode)
71
72   Open an audio device and return an OSS audio device object.  This object
73   supports many file-like methods, such as :meth:`read`, :meth:`write`, and
74   :meth:`fileno` (although there are subtle differences between conventional Unix
75   read/write semantics and those of OSS audio devices).  It also supports a number
76   of audio-specific methods; see below for the complete list of methods.
77
78   *device* is the audio device filename to use.  If it is not specified, this
79   module first looks in the environment variable :envvar:`AUDIODEV` for a device
80   to use.  If not found, it falls back to :file:`/dev/dsp`.
81
82   *mode* is one of ``'r'`` for read-only (record) access, ``'w'`` for
83   write-only (playback) access and ``'rw'`` for both. Since many sound cards
84   only allow one process to have the recorder or player open at a time, it is a
85   good idea to open the device only for the activity needed.  Further, some
86   sound cards are half-duplex: they can be opened for reading or writing, but
87   not both at once.
88
89   Note the unusual calling syntax: the *first* argument is optional, and the
90   second is required.  This is a historical artifact for compatibility with the
91   older :mod:`linuxaudiodev` module which :mod:`ossaudiodev` supersedes.
92
93   .. XXX it might also be motivated
94      by my unfounded-but-still-possibly-true belief that the default
95      audio device varies unpredictably across operating systems.  -GW
96
97
98.. function:: openmixer([device])
99
100   Open a mixer device and return an OSS mixer device object.   *device* is the
101   mixer device filename to use.  If it is not specified, this module first looks
102   in the environment variable :envvar:`MIXERDEV` for a device to use.  If not
103   found, it falls back to :file:`/dev/mixer`.
104
105
106.. _ossaudio-device-objects:
107
108Audio Device Objects
109--------------------
110
111Before you can write to or read from an audio device, you must call three
112methods in the correct order:
113
114#. :meth:`setfmt` to set the output format
115
116#. :meth:`channels` to set the number of channels
117
118#. :meth:`speed` to set the sample rate
119
120Alternately, you can use the :meth:`setparameters` method to set all three audio
121parameters at once.  This is more convenient, but may not be as flexible in all
122cases.
123
124The audio device objects returned by :func:`.open` define the following methods
125and (read-only) attributes:
126
127
128.. method:: oss_audio_device.close()
129
130   Explicitly close the audio device.  When you are done writing to or reading from
131   an audio device, you should explicitly close it.  A closed device cannot be used
132   again.
133
134
135.. method:: oss_audio_device.fileno()
136
137   Return the file descriptor associated with the device.
138
139
140.. method:: oss_audio_device.read(size)
141
142   Read *size* bytes from the audio input and return them as a Python string.
143   Unlike most Unix device drivers, OSS audio devices in blocking mode (the
144   default) will block :func:`read` until the entire requested amount of data is
145   available.
146
147
148.. method:: oss_audio_device.write(data)
149
150   Write the Python string *data* to the audio device and return the number of
151   bytes written.  If the audio device is in blocking mode (the default), the
152   entire string is always written (again, this is different from usual Unix device
153   semantics).  If the device is in non-blocking mode, some data may not be written
154   ---see :meth:`writeall`.
155
156
157.. method:: oss_audio_device.writeall(data)
158
159   Write the entire Python string *data* to the audio device: waits until the audio
160   device is able to accept data, writes as much data as it will accept, and
161   repeats until *data* has been completely written. If the device is in blocking
162   mode (the default), this has the same effect as :meth:`write`; :meth:`writeall`
163   is only useful in non-blocking mode.  Has no return value, since the amount of
164   data written is always equal to the amount of data supplied.
165
166The following methods each map to exactly one :c:func:`ioctl` system call.  The
167correspondence is obvious: for example, :meth:`setfmt` corresponds to the
168``SNDCTL_DSP_SETFMT`` ioctl, and :meth:`sync` to ``SNDCTL_DSP_SYNC`` (this can
169be useful when consulting the OSS documentation).  If the underlying
170:c:func:`ioctl` fails, they all raise :exc:`IOError`.
171
172
173.. method:: oss_audio_device.nonblock()
174
175   Put the device into non-blocking mode.  Once in non-blocking mode, there is no
176   way to return it to blocking mode.
177
178
179.. method:: oss_audio_device.getfmts()
180
181   Return a bitmask of the audio output formats supported by the soundcard.  Some
182   of the formats supported by OSS are:
183
184   +-------------------------+---------------------------------------------+
185   | Format                  | Description                                 |
186   +=========================+=============================================+
187   | :const:`AFMT_MU_LAW`    | a logarithmic encoding (used by Sun ``.au`` |
188   |                         | files and :file:`/dev/audio`)               |
189   +-------------------------+---------------------------------------------+
190   | :const:`AFMT_A_LAW`     | a logarithmic encoding                      |
191   +-------------------------+---------------------------------------------+
192   | :const:`AFMT_IMA_ADPCM` | a 4:1 compressed format defined by the      |
193   |                         | Interactive Multimedia Association          |
194   +-------------------------+---------------------------------------------+
195   | :const:`AFMT_U8`        | Unsigned, 8-bit audio                       |
196   +-------------------------+---------------------------------------------+
197   | :const:`AFMT_S16_LE`    | Signed, 16-bit audio, little-endian byte    |
198   |                         | order (as used by Intel processors)         |
199   +-------------------------+---------------------------------------------+
200   | :const:`AFMT_S16_BE`    | Signed, 16-bit audio, big-endian byte order |
201   |                         | (as used by 68k, PowerPC, Sparc)            |
202   +-------------------------+---------------------------------------------+
203   | :const:`AFMT_S8`        | Signed, 8 bit audio                         |
204   +-------------------------+---------------------------------------------+
205   | :const:`AFMT_U16_LE`    | Unsigned, 16-bit little-endian audio        |
206   +-------------------------+---------------------------------------------+
207   | :const:`AFMT_U16_BE`    | Unsigned, 16-bit big-endian audio           |
208   +-------------------------+---------------------------------------------+
209
210   Consult the OSS documentation for a full list of audio formats, and note that
211   most devices support only a subset of these formats.  Some older devices only
212   support :const:`AFMT_U8`; the most common format used today is
213   :const:`AFMT_S16_LE`.
214
215
216.. method:: oss_audio_device.setfmt(format)
217
218   Try to set the current audio format to *format*---see :meth:`getfmts` for a
219   list.  Returns the audio format that the device was set to, which may not be the
220   requested format.  May also be used to return the current audio format---do this
221   by passing an "audio format" of :const:`AFMT_QUERY`.
222
223
224.. method:: oss_audio_device.channels(nchannels)
225
226   Set the number of output channels to *nchannels*.  A value of 1 indicates
227   monophonic sound, 2 stereophonic.  Some devices may have more than 2 channels,
228   and some high-end devices may not support mono. Returns the number of channels
229   the device was set to.
230
231
232.. method:: oss_audio_device.speed(samplerate)
233
234   Try to set the audio sampling rate to *samplerate* samples per second.  Returns
235   the rate actually set.  Most sound devices don't support arbitrary sampling
236   rates.  Common rates are:
237
238   +-------+-------------------------------------------+
239   | Rate  | Description                               |
240   +=======+===========================================+
241   | 8000  | default rate for :file:`/dev/audio`       |
242   +-------+-------------------------------------------+
243   | 11025 | speech recording                          |
244   +-------+-------------------------------------------+
245   | 22050 |                                           |
246   +-------+-------------------------------------------+
247   | 44100 | CD quality audio (at 16 bits/sample and 2 |
248   |       | channels)                                 |
249   +-------+-------------------------------------------+
250   | 96000 | DVD quality audio (at 24 bits/sample)     |
251   +-------+-------------------------------------------+
252
253
254.. method:: oss_audio_device.sync()
255
256   Wait until the sound device has played every byte in its buffer.  (This happens
257   implicitly when the device is closed.)  The OSS documentation recommends closing
258   and re-opening the device rather than using :meth:`sync`.
259
260
261.. method:: oss_audio_device.reset()
262
263   Immediately stop playing or recording and return the device to a state where it
264   can accept commands.  The OSS documentation recommends closing and re-opening
265   the device after calling :meth:`reset`.
266
267
268.. method:: oss_audio_device.post()
269
270   Tell the driver that there is likely to be a pause in the output, making it
271   possible for the device to handle the pause more intelligently.  You might use
272   this after playing a spot sound effect, before waiting for user input, or before
273   doing disk I/O.
274
275The following convenience methods combine several ioctls, or one ioctl and some
276simple calculations.
277
278
279.. method:: oss_audio_device.setparameters(format, nchannels, samplerate[, strict=False])
280
281   Set the key audio sampling parameters---sample format, number of channels, and
282   sampling rate---in one method call.  *format*,  *nchannels*, and *samplerate*
283   should be as specified in the :meth:`setfmt`, :meth:`channels`, and
284   :meth:`speed`  methods.  If *strict* is true, :meth:`setparameters` checks to
285   see if each parameter was actually set to the requested value, and raises
286   :exc:`OSSAudioError` if not.  Returns a tuple (*format*, *nchannels*,
287   *samplerate*) indicating the parameter values that were actually set by the
288   device driver (i.e., the same as the return values of :meth:`setfmt`,
289   :meth:`channels`, and :meth:`speed`).
290
291   For example,  ::
292
293      (fmt, channels, rate) = dsp.setparameters(fmt, channels, rate)
294
295   is equivalent to  ::
296
297      fmt = dsp.setfmt(fmt)
298      channels = dsp.channels(channels)
299      rate = dsp.rate(rate)
300
301
302.. method:: oss_audio_device.bufsize()
303
304   Returns the size of the hardware buffer, in samples.
305
306
307.. method:: oss_audio_device.obufcount()
308
309   Returns the number of samples that are in the hardware buffer yet to be played.
310
311
312.. method:: oss_audio_device.obuffree()
313
314   Returns the number of samples that could be queued into the hardware buffer to
315   be played without blocking.
316
317Audio device objects also support several read-only attributes:
318
319
320.. attribute:: oss_audio_device.closed
321
322   Boolean indicating whether the device has been closed.
323
324
325.. attribute:: oss_audio_device.name
326
327   String containing the name of the device file.
328
329
330.. attribute:: oss_audio_device.mode
331
332   The I/O mode for the file, either ``"r"``, ``"rw"``, or ``"w"``.
333
334
335.. _mixer-device-objects:
336
337Mixer Device Objects
338--------------------
339
340The mixer object provides two file-like methods:
341
342
343.. method:: oss_mixer_device.close()
344
345   This method closes the open mixer device file.  Any further attempts to use the
346   mixer after this file is closed will raise an :exc:`IOError`.
347
348
349.. method:: oss_mixer_device.fileno()
350
351   Returns the file handle number of the open mixer device file.
352
353The remaining methods are specific to audio mixing:
354
355
356.. method:: oss_mixer_device.controls()
357
358   This method returns a bitmask specifying the available mixer controls ("Control"
359   being a specific mixable "channel", such as :const:`SOUND_MIXER_PCM` or
360   :const:`SOUND_MIXER_SYNTH`).  This bitmask indicates a subset of all available
361   mixer controls---the :const:`SOUND_MIXER_\*` constants defined at module level.
362   To determine if, for example, the current mixer object supports a PCM mixer, use
363   the following Python code::
364
365      mixer=ossaudiodev.openmixer()
366      if mixer.controls() & (1 << ossaudiodev.SOUND_MIXER_PCM):
367          # PCM is supported
368          ... code ...
369
370   For most purposes, the :const:`SOUND_MIXER_VOLUME` (master volume) and
371   :const:`SOUND_MIXER_PCM` controls should suffice---but code that uses the mixer
372   should be flexible when it comes to choosing mixer controls.  On the Gravis
373   Ultrasound, for example, :const:`SOUND_MIXER_VOLUME` does not exist.
374
375
376.. method:: oss_mixer_device.stereocontrols()
377
378   Returns a bitmask indicating stereo mixer controls.  If a bit is set, the
379   corresponding control is stereo; if it is unset, the control is either
380   monophonic or not supported by the mixer (use in combination with
381   :meth:`controls` to determine which).
382
383   See the code example for the :meth:`controls` function for an example of getting
384   data from a bitmask.
385
386
387.. method:: oss_mixer_device.reccontrols()
388
389   Returns a bitmask specifying the mixer controls that may be used to record.  See
390   the code example for :meth:`controls` for an example of reading from a bitmask.
391
392
393.. method:: oss_mixer_device.get(control)
394
395   Returns the volume of a given mixer control.  The returned volume is a 2-tuple
396   ``(left_volume,right_volume)``.  Volumes are specified as numbers from 0
397   (silent) to 100 (full volume).  If the control is monophonic, a 2-tuple is still
398   returned, but both volumes are the same.
399
400   Raises :exc:`OSSAudioError` if an invalid control was is specified, or
401   :exc:`IOError` if an unsupported control is specified.
402
403
404.. method:: oss_mixer_device.set(control, (left, right))
405
406   Sets the volume for a given mixer control to ``(left,right)``. ``left`` and
407   ``right`` must be ints and between 0 (silent) and 100 (full volume).  On
408   success, the new volume is returned as a 2-tuple. Note that this may not be
409   exactly the same as the volume specified, because of the limited resolution of
410   some soundcard's mixers.
411
412   Raises :exc:`OSSAudioError` if an invalid mixer control was specified, or if the
413   specified volumes were out-of-range.
414
415
416.. method:: oss_mixer_device.get_recsrc()
417
418   This method returns a bitmask indicating which control(s) are currently being
419   used as a recording source.
420
421
422.. method:: oss_mixer_device.set_recsrc(bitmask)
423
424   Call this function to specify a recording source.  Returns a bitmask indicating
425   the new recording source (or sources) if successful; raises :exc:`IOError` if an
426   invalid source was specified.  To set the current recording source to the
427   microphone input::
428
429      mixer.setrecsrc (1 << ossaudiodev.SOUND_MIXER_MIC)
430
431