Ei vielä käännetty
Tätä sivua ei ole vielä käännetty suomeksi, joten se näytetään englanniksi. Auta kääntämään
ossaudiodev Module¶
⚠️ REMOVED IN PYTHON 3.13: The ossaudiodev module was deprecated in Python 3.11 and removed in Python 3.13.
The ossaudiodev module provides access to OSS (Open Sound System) audio
devices on Linux and FreeBSD: opening /dev/dsp for playback or recording, and
/dev/mixer for volume control.
Throughout, n is the number of bytes transferred.
Complexity Reference¶
Module¶
| Operation | Time | Space | Notes |
|---|---|---|---|
open(mode) |
O(1) + syscall | O(1) | Opens /dev/dsp; mode is "r", "w", or "rw" |
open(device, mode) |
O(1) + syscall | O(1) | Explicit device path |
openmixer(device=None) |
O(1) + syscall | O(1) | Opens /dev/mixer |
Audio Device¶
| Operation | Time | Space | Notes |
|---|---|---|---|
read(size) |
O(n) + blocks | O(n) | Blocks until size bytes are captured |
write(data) |
O(n) + blocks | O(1) | Blocks until the buffer accepts the data |
writeall(data) |
O(n) + blocks | O(1) | Loops until everything is written |
setparameters(format, nchannels, samplerate) |
O(1) + syscall | O(1) | Configure the device |
setfmt(format) |
O(1) + syscall | O(1) | Sample format only |
channels(n) |
O(1) + syscall | O(1) | Channel count only |
speed(rate) |
O(1) + syscall | O(1) | Sample rate only |
bufsize() |
O(1) + syscall | O(1) | Buffer size in samples |
obufcount() |
O(1) + syscall | O(1) | Samples queued for playback |
obuffree() |
O(1) + syscall | O(1) | Space remaining in the buffer |
sync() |
O(1) + blocks | O(1) | Waits for the buffer to drain |
reset() |
O(1) + syscall | O(1) | Discards buffered audio |
close() |
O(1) + syscall | O(1) | Release the device |
Mixer Device¶
| Operation | Time | Space | Notes |
|---|---|---|---|
controls() |
O(1) + syscall | O(1) | Bitmask of supported controls |
get(control) |
O(1) + syscall | O(1) | Current volume |
set(control, (left, right)) |
O(1) + syscall | O(1) | Set volume |
stereocontrols() |
O(1) + syscall | O(1) | Which controls are stereo |
The data-transfer calls are linear in bytes; everything else is a constant-time
ioctl. The dominant cost in practice is blocking on the device, not CPU work.
Playing Audio¶
import ossaudiodev
# O(1) to open and configure
dsp = ossaudiodev.open("w")
try:
dsp.setparameters(ossaudiodev.AFMT_S16_LE, 2, 44100)
dsp.writeall(pcm_bytes) # O(n), blocks until fully written
dsp.sync() # blocks until the buffer drains
finally:
dsp.close()
Recording Audio¶
import ossaudiodev
dsp = ossaudiodev.open("r")
try:
dsp.setparameters(ossaudiodev.AFMT_S16_LE, 1, 16000)
# O(n) per chunk; blocks until the samples are captured
chunks = [dsp.read(4096) for _ in range(100)]
finally:
dsp.close()
audio = b"".join(chunks) # O(total bytes)
Avoiding Underruns¶
write() blocks when the device buffer is full. To keep a real-time loop
responsive, check the free space first:
import ossaudiodev
dsp = ossaudiodev.open("w")
dsp.setparameters(ossaudiodev.AFMT_S16_LE, 2, 44100)
# O(1) - only write what fits, never block
free = dsp.obuffree()
if free >= len(chunk):
dsp.write(chunk)
Removed in Python 3.13
There is no standard-library replacement. OSS itself is largely superseded by ALSA and PulseAudio on Linux; use a third-party binding for those.
Linux and FreeBSD only
The module was never available on Windows or macOS, and requires the OSS device nodes to be present.
Version Notes¶
- Python 3.11: deprecated (PEP 594)
- Python 3.13: removed
- Before 3.13: available only on platforms with OSS device support
- All versions: transfer calls are linear in bytes; control calls are
constant-time
ioctls