-
Notifications
You must be signed in to change notification settings - Fork 112
/
soundfile.py
1674 lines (1432 loc) · 63.3 KB
/
soundfile.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
"""python-soundfile is an audio library based on libsndfile, CFFI and NumPy.
Sound files can be read or written directly using the functions
`read()` and `write()`.
To read a sound file in a block-wise fashion, use `blocks()`.
Alternatively, sound files can be opened as `SoundFile` objects.
For further information, see https://python-soundfile.readthedocs.io/.
"""
__version__ = "0.13.0"
import os as _os
import sys as _sys
from os import SEEK_SET, SEEK_CUR, SEEK_END
from ctypes.util import find_library as _find_library
from _soundfile import ffi as _ffi
try:
_unicode = unicode # doesn't exist in Python 3.x
except NameError:
_unicode = str
_str_types = {
'title': 0x01,
'copyright': 0x02,
'software': 0x03,
'artist': 0x04,
'comment': 0x05,
'date': 0x06,
'album': 0x07,
'license': 0x08,
'tracknumber': 0x09,
'genre': 0x10,
}
_formats = {
'WAV': 0x010000, # Microsoft WAV format (little endian default).
'AIFF': 0x020000, # Apple/SGI AIFF format (big endian).
'AU': 0x030000, # Sun/NeXT AU format (big endian).
'RAW': 0x040000, # RAW PCM data.
'PAF': 0x050000, # Ensoniq PARIS file format.
'SVX': 0x060000, # Amiga IFF / SVX8 / SV16 format.
'NIST': 0x070000, # Sphere NIST format.
'VOC': 0x080000, # VOC files.
'IRCAM': 0x0A0000, # Berkeley/IRCAM/CARL
'W64': 0x0B0000, # Sonic Foundry's 64 bit RIFF/WAV
'MAT4': 0x0C0000, # Matlab (tm) V4.2 / GNU Octave 2.0
'MAT5': 0x0D0000, # Matlab (tm) V5.0 / GNU Octave 2.1
'PVF': 0x0E0000, # Portable Voice Format
'XI': 0x0F0000, # Fasttracker 2 Extended Instrument
'HTK': 0x100000, # HMM Tool Kit format
'SDS': 0x110000, # Midi Sample Dump Standard
'AVR': 0x120000, # Audio Visual Research
'WAVEX': 0x130000, # MS WAVE with WAVEFORMATEX
'SD2': 0x160000, # Sound Designer 2
'FLAC': 0x170000, # FLAC lossless file format
'CAF': 0x180000, # Core Audio File format
'WVE': 0x190000, # Psion WVE format
'OGG': 0x200000, # Xiph OGG container
'MPC2K': 0x210000, # Akai MPC 2000 sampler
'RF64': 0x220000, # RF64 WAV file
'MP3': 0x230000, # MPEG-1/2 audio stream
}
_subtypes = {
'PCM_S8': 0x0001, # Signed 8 bit data
'PCM_16': 0x0002, # Signed 16 bit data
'PCM_24': 0x0003, # Signed 24 bit data
'PCM_32': 0x0004, # Signed 32 bit data
'PCM_U8': 0x0005, # Unsigned 8 bit data (WAV and RAW only)
'FLOAT': 0x0006, # 32 bit float data
'DOUBLE': 0x0007, # 64 bit float data
'ULAW': 0x0010, # U-Law encoded.
'ALAW': 0x0011, # A-Law encoded.
'IMA_ADPCM': 0x0012, # IMA ADPCM.
'MS_ADPCM': 0x0013, # Microsoft ADPCM.
'GSM610': 0x0020, # GSM 6.10 encoding.
'VOX_ADPCM': 0x0021, # OKI / Dialogix ADPCM
'NMS_ADPCM_16': 0x0022, # 16kbs NMS G721-variant encoding.
'NMS_ADPCM_24': 0x0023, # 24kbs NMS G721-variant encoding.
'NMS_ADPCM_32': 0x0024, # 32kbs NMS G721-variant encoding.
'G721_32': 0x0030, # 32kbs G721 ADPCM encoding.
'G723_24': 0x0031, # 24kbs G723 ADPCM encoding.
'G723_40': 0x0032, # 40kbs G723 ADPCM encoding.
'DWVW_12': 0x0040, # 12 bit Delta Width Variable Word encoding.
'DWVW_16': 0x0041, # 16 bit Delta Width Variable Word encoding.
'DWVW_24': 0x0042, # 24 bit Delta Width Variable Word encoding.
'DWVW_N': 0x0043, # N bit Delta Width Variable Word encoding.
'DPCM_8': 0x0050, # 8 bit differential PCM (XI only)
'DPCM_16': 0x0051, # 16 bit differential PCM (XI only)
'VORBIS': 0x0060, # Xiph Vorbis encoding.
'OPUS': 0x0064, # Xiph/Skype Opus encoding.
'ALAC_16': 0x0070, # Apple Lossless Audio Codec (16 bit).
'ALAC_20': 0x0071, # Apple Lossless Audio Codec (20 bit).
'ALAC_24': 0x0072, # Apple Lossless Audio Codec (24 bit).
'ALAC_32': 0x0073, # Apple Lossless Audio Codec (32 bit).
'MPEG_LAYER_I': 0x0080, # MPEG-1 Audio Layer I.
'MPEG_LAYER_II': 0x0081, # MPEG-1 Audio Layer II.
'MPEG_LAYER_III': 0x0082, # MPEG-2 Audio Layer III.
}
_endians = {
'FILE': 0x00000000, # Default file endian-ness.
'LITTLE': 0x10000000, # Force little endian-ness.
'BIG': 0x20000000, # Force big endian-ness.
'CPU': 0x30000000, # Force CPU endian-ness.
}
# libsndfile doesn't specify default subtypes, these are somehow arbitrary:
_default_subtypes = {
'WAV': 'PCM_16',
'AIFF': 'PCM_16',
'AU': 'PCM_16',
# 'RAW': # subtype must be explicit!
'PAF': 'PCM_16',
'SVX': 'PCM_16',
'NIST': 'PCM_16',
'VOC': 'PCM_16',
'IRCAM': 'PCM_16',
'W64': 'PCM_16',
'MAT4': 'DOUBLE',
'MAT5': 'DOUBLE',
'PVF': 'PCM_16',
'XI': 'DPCM_16',
'HTK': 'PCM_16',
'SDS': 'PCM_16',
'AVR': 'PCM_16',
'WAVEX': 'PCM_16',
'SD2': 'PCM_16',
'FLAC': 'PCM_16',
'CAF': 'PCM_16',
'WVE': 'ALAW',
'OGG': 'VORBIS',
'MPC2K': 'PCM_16',
'RF64': 'PCM_16',
'MP3': 'MPEG_LAYER_III',
}
_ffi_types = {
'float64': 'double',
'float32': 'float',
'int32': 'int',
'int16': 'short'
}
_bitrate_modes = {
'CONSTANT': 0,
'AVERAGE': 1,
'VARIABLE': 2,
}
try: # packaged lib (in _soundfile_data which should be on python path)
if _sys.platform == 'darwin':
from platform import machine as _machine
_packaged_libname = 'libsndfile_' + _machine() + '.dylib'
elif _sys.platform == 'win32':
from platform import architecture as _architecture
from platform import machine as _machine
# this check can not be completed correctly: for x64 binaries running on
# arm64 Windows report the same values as arm64 binaries. For now, neither
# numpy nor cffi are available for arm64, so we can safely assume we're
# in x86 land:
if _architecture()[0] == '64bit':
_packaged_libname = 'libsndfile_x64.dll'
elif _architecture()[0] == '32bit':
_packaged_libname = 'libsndfile_x86.dll'
else:
raise OSError('no packaged library for Windows {} {}'
.format(_architecture(), _machine()))
elif _sys.platform == 'linux':
from platform import machine as _machine
if _machine() in ["aarch64", "aarch64_be", "armv8b", "armv8l"]:
_packaged_libname = 'libsndfile_arm64.so'
else:
_packaged_libname = 'libsndfile_' + _machine() + '.so'
else:
raise OSError('no packaged library for this platform')
import _soundfile_data # ImportError if this doesn't exist
_path = _os.path.dirname(_soundfile_data.__file__) # TypeError if __file__ is None
_full_path = _os.path.join(_path, _packaged_libname)
_snd = _ffi.dlopen(_full_path) # OSError if file doesn't exist or can't be loaded
except (OSError, ImportError, TypeError):
try: # system-wide libsndfile:
_libname = _find_library('sndfile')
if _libname is None:
raise OSError('sndfile library not found using ctypes.util.find_library')
_snd = _ffi.dlopen(_libname)
except OSError:
# Try explicit file name, if the general does not work (e.g. on nixos)
if _sys.platform == 'darwin':
_explicit_libname = 'libsndfile.dylib'
elif _sys.platform == 'win32':
_explicit_libname = 'libsndfile.dll'
elif _sys.platform == 'linux':
_explicit_libname = 'libsndfile.so'
else:
raise
# Homebrew on Apple M1 uses a `/opt/homebrew/lib` instead of
# `/usr/local/lib`. We are making sure we pick that up.
from platform import machine as _machine
if _sys.platform == 'darwin' and _machine() == 'arm64':
_hbrew_path = '/opt/homebrew/lib/' if _os.path.isdir('/opt/homebrew/lib/') \
else '/usr/local/lib/'
_snd = _ffi.dlopen(_os.path.join(_hbrew_path, _explicit_libname))
else:
_snd = _ffi.dlopen(_explicit_libname)
__libsndfile_version__ = _ffi.string(_snd.sf_version_string()).decode('utf-8', 'replace')
if __libsndfile_version__.startswith('libsndfile-'):
__libsndfile_version__ = __libsndfile_version__[len('libsndfile-'):]
def read(file, frames=-1, start=0, stop=None, dtype='float64', always_2d=False,
fill_value=None, out=None, samplerate=None, channels=None,
format=None, subtype=None, endian=None, closefd=True):
"""Provide audio data from a sound file as NumPy array.
By default, the whole file is read from the beginning, but the
position to start reading can be specified with *start* and the
number of frames to read can be specified with *frames*.
Alternatively, a range can be specified with *start* and *stop*.
If there is less data left in the file than requested, the rest of
the frames are filled with *fill_value*.
If no *fill_value* is specified, a smaller array is returned.
Parameters
----------
file : str or int or file-like object
The file to read from. See `SoundFile` for details.
frames : int, optional
The number of frames to read. If *frames* is negative, the whole
rest of the file is read. Not allowed if *stop* is given.
start : int, optional
Where to start reading. A negative value counts from the end.
stop : int, optional
The index after the last frame to be read. A negative value
counts from the end. Not allowed if *frames* is given.
dtype : {'float64', 'float32', 'int32', 'int16'}, optional
Data type of the returned array, by default ``'float64'``.
Floating point audio data is typically in the range from
``-1.0`` to ``1.0``. Integer data is in the range from
``-2**15`` to ``2**15-1`` for ``'int16'`` and from ``-2**31`` to
``2**31-1`` for ``'int32'``.
.. note:: Reading int values from a float file will *not*
scale the data to [-1.0, 1.0). If the file contains
``np.array([42.6], dtype='float32')``, you will read
``np.array([43], dtype='int32')`` for ``dtype='int32'``.
Returns
-------
audiodata : `numpy.ndarray` or type(out)
A two-dimensional (frames x channels) NumPy array is returned.
If the sound file has only one channel, a one-dimensional array
is returned. Use ``always_2d=True`` to return a two-dimensional
array anyway.
If *out* was specified, it is returned. If *out* has more
frames than available in the file (or if *frames* is smaller
than the length of *out*) and no *fill_value* is given, then
only a part of *out* is overwritten and a view containing all
valid frames is returned.
samplerate : int
The sample rate of the audio file.
Other Parameters
----------------
always_2d : bool, optional
By default, reading a mono sound file will return a
one-dimensional array. With ``always_2d=True``, audio data is
always returned as a two-dimensional array, even if the audio
file has only one channel.
fill_value : float, optional
If more frames are requested than available in the file, the
rest of the output is be filled with *fill_value*. If
*fill_value* is not specified, a smaller array is returned.
out : `numpy.ndarray` or subclass, optional
If *out* is specified, the data is written into the given array
instead of creating a new array. In this case, the arguments
*dtype* and *always_2d* are silently ignored! If *frames* is
not given, it is obtained from the length of *out*.
samplerate, channels, format, subtype, endian, closefd
See `SoundFile`.
Examples
--------
>>> import soundfile as sf
>>> data, samplerate = sf.read('stereo_file.wav')
>>> data
array([[ 0.71329652, 0.06294799],
[-0.26450912, -0.38874483],
...
[ 0.67398441, -0.11516333]])
>>> samplerate
44100
"""
with SoundFile(file, 'r', samplerate, channels,
subtype, endian, format, closefd) as f:
frames = f._prepare_read(start, stop, frames)
data = f.read(frames, dtype, always_2d, fill_value, out)
return data, f.samplerate
def write(file, data, samplerate, subtype=None, endian=None, format=None,
closefd=True, compression_level=None, bitrate_mode=None):
"""Write data to a sound file.
.. note:: If *file* exists, it will be truncated and overwritten!
Parameters
----------
file : str or int or file-like object
The file to write to. See `SoundFile` for details.
data : array_like
The data to write. Usually two-dimensional (frames x channels),
but one-dimensional *data* can be used for mono files.
Only the data types ``'float64'``, ``'float32'``, ``'int32'``
and ``'int16'`` are supported.
.. note:: The data type of *data* does **not** select the data
type of the written file. Audio data will be
converted to the given *subtype*. Writing int values
to a float file will *not* scale the values to
[-1.0, 1.0). If you write the value ``np.array([42],
dtype='int32')``, to a ``subtype='FLOAT'`` file, the
file will then contain ``np.array([42.],
dtype='float32')``.
samplerate : int
The sample rate of the audio data.
subtype : str, optional
See `default_subtype()` for the default value and
`available_subtypes()` for all possible values.
Other Parameters
----------------
format, endian, closefd, compression_level, bitrate_mode
See `SoundFile`.
Examples
--------
Write 10 frames of random data to a new file:
>>> import numpy as np
>>> import soundfile as sf
>>> sf.write('stereo_file.wav', np.random.randn(10, 2), 44100, 'PCM_24')
"""
import numpy as np
data = np.asarray(data)
if data.ndim == 1:
channels = 1
else:
channels = data.shape[1]
with SoundFile(file, 'w', samplerate, channels,
subtype, endian, format, closefd,
compression_level, bitrate_mode) as f:
f.write(data)
def blocks(file, blocksize=None, overlap=0, frames=-1, start=0, stop=None,
dtype='float64', always_2d=False, fill_value=None, out=None,
samplerate=None, channels=None,
format=None, subtype=None, endian=None, closefd=True):
"""Return a generator for block-wise reading.
By default, iteration starts at the beginning and stops at the end
of the file. Use *start* to start at a later position and *frames*
or *stop* to stop earlier.
If you stop iterating over the generator before it's exhausted,
the sound file is not closed. This is normally not a problem
because the file is opened in read-only mode. To close the file
properly, the generator's ``close()`` method can be called.
Parameters
----------
file : str or int or file-like object
The file to read from. See `SoundFile` for details.
blocksize : int
The number of frames to read per block.
Either this or *out* must be given.
overlap : int, optional
The number of frames to rewind between each block.
Yields
------
`numpy.ndarray` or type(out)
Blocks of audio data.
If *out* was given, and the requested frames are not an integer
multiple of the length of *out*, and no *fill_value* was given,
the last block will be a smaller view into *out*.
Other Parameters
----------------
frames, start, stop
See `read()`.
dtype : {'float64', 'float32', 'int32', 'int16'}, optional
See `read()`.
always_2d, fill_value, out
See `read()`.
samplerate, channels, format, subtype, endian, closefd
See `SoundFile`.
Examples
--------
>>> import soundfile as sf
>>> for block in sf.blocks('stereo_file.wav', blocksize=1024):
>>> pass # do something with 'block'
"""
with SoundFile(file, 'r', samplerate, channels,
subtype, endian, format, closefd) as f:
frames = f._prepare_read(start, stop, frames)
for block in f.blocks(blocksize, overlap, frames,
dtype, always_2d, fill_value, out):
yield block
class _SoundFileInfo(object):
"""Information about a SoundFile"""
def __init__(self, file, verbose):
self.verbose = verbose
with SoundFile(file) as f:
self.name = f.name
self.samplerate = f.samplerate
self.channels = f.channels
self.frames = f.frames
self.duration = float(self.frames)/f.samplerate
self.format = f.format
self.subtype = f.subtype
self.endian = f.endian
self.format_info = f.format_info
self.subtype_info = f.subtype_info
self.sections = f.sections
self.extra_info = f.extra_info
@property
def _duration_str(self):
hours, rest = divmod(self.duration, 3600)
minutes, seconds = divmod(rest, 60)
if hours >= 1:
duration = "{0:.0g}:{1:02.0g}:{2:05.3f} h".format(hours, minutes, seconds)
elif minutes >= 1:
duration = "{0:02.0g}:{1:05.3f} min".format(minutes, seconds)
elif seconds <= 1:
duration = "{0:d} samples".format(self.frames)
else:
duration = "{0:.3f} s".format(seconds)
return duration
def __repr__(self):
info = "\n".join(
["{0.name}",
"samplerate: {0.samplerate} Hz",
"channels: {0.channels}",
"duration: {0._duration_str}",
"format: {0.format_info} [{0.format}]",
"subtype: {0.subtype_info} [{0.subtype}]"])
if self.verbose:
info += "\n".join(
["\nendian: {0.endian}",
"sections: {0.sections}",
"frames: {0.frames}",
'extra_info: """',
' {1}"""'])
indented_extra_info = ("\n"+" "*4).join(self.extra_info.split("\n"))
return info.format(self, indented_extra_info)
def info(file, verbose=False):
"""Returns an object with information about a `SoundFile`.
Parameters
----------
verbose : bool
Whether to print additional information.
"""
return _SoundFileInfo(file, verbose)
def available_formats():
"""Return a dictionary of available major formats.
Examples
--------
>>> import soundfile as sf
>>> sf.available_formats()
{'FLAC': 'FLAC (FLAC Lossless Audio Codec)',
'OGG': 'OGG (OGG Container format)',
'WAV': 'WAV (Microsoft)',
'AIFF': 'AIFF (Apple/SGI)',
...
'WAVEX': 'WAVEX (Microsoft)',
'RAW': 'RAW (header-less)',
'MAT5': 'MAT5 (GNU Octave 2.1 / Matlab 5.0)'}
"""
return dict(_available_formats_helper(_snd.SFC_GET_FORMAT_MAJOR_COUNT,
_snd.SFC_GET_FORMAT_MAJOR))
def available_subtypes(format=None):
"""Return a dictionary of available subtypes.
Parameters
----------
format : str
If given, only compatible subtypes are returned.
Examples
--------
>>> import soundfile as sf
>>> sf.available_subtypes('FLAC')
{'PCM_24': 'Signed 24 bit PCM',
'PCM_16': 'Signed 16 bit PCM',
'PCM_S8': 'Signed 8 bit PCM'}
"""
subtypes = _available_formats_helper(_snd.SFC_GET_FORMAT_SUBTYPE_COUNT,
_snd.SFC_GET_FORMAT_SUBTYPE)
return dict((subtype, name) for subtype, name in subtypes
if format is None or check_format(format, subtype))
def check_format(format, subtype=None, endian=None):
"""Check if the combination of format/subtype/endian is valid.
Examples
--------
>>> import soundfile as sf
>>> sf.check_format('WAV', 'PCM_24')
True
>>> sf.check_format('FLAC', 'VORBIS')
False
"""
try:
return bool(_format_int(format, subtype, endian))
except (ValueError, TypeError):
return False
def default_subtype(format):
"""Return the default subtype for a given format.
Examples
--------
>>> import soundfile as sf
>>> sf.default_subtype('WAV')
'PCM_16'
>>> sf.default_subtype('MAT5')
'DOUBLE'
"""
_check_format(format)
return _default_subtypes.get(format.upper())
class SoundFile(object):
"""A sound file.
For more documentation see the __init__() docstring (which is also
used for the online documentation (https://python-soundfile.readthedocs.io/).
"""
def __init__(self, file, mode='r', samplerate=None, channels=None,
subtype=None, endian=None, format=None, closefd=True,
compression_level=None, bitrate_mode=None):
"""Open a sound file.
If a file is opened with `mode` ``'r'`` (the default) or
``'r+'``, no sample rate, channels or file format need to be
given because the information is obtained from the file. An
exception is the ``'RAW'`` data format, which always requires
these data points.
File formats consist of three case-insensitive strings:
* a *major format* which is by default obtained from the
extension of the file name (if known) and which can be
forced with the format argument (e.g. ``format='WAVEX'``).
* a *subtype*, e.g. ``'PCM_24'``. Most major formats have a
default subtype which is used if no subtype is specified.
* an *endian-ness*, which doesn't have to be specified at all in
most cases.
A `SoundFile` object is a *context manager*, which means
if used in a "with" statement, `close()` is automatically
called when reaching the end of the code block inside the "with"
statement.
Parameters
----------
file : str or int or file-like object
The file to open. This can be a file name, a file
descriptor or a Python file object (or a similar object with
the methods ``read()``/``readinto()``, ``write()``,
``seek()`` and ``tell()``).
mode : {'r', 'r+', 'w', 'w+', 'x', 'x+'}, optional
Open mode. Has to begin with one of these three characters:
``'r'`` for reading, ``'w'`` for writing (truncates *file*)
or ``'x'`` for writing (raises an error if *file* already
exists). Additionally, it may contain ``'+'`` to open
*file* for both reading and writing.
The character ``'b'`` for *binary mode* is implied because
all sound files have to be opened in this mode.
If *file* is a file descriptor or a file-like object,
``'w'`` doesn't truncate and ``'x'`` doesn't raise an error.
samplerate : int
The sample rate of the file. If `mode` contains ``'r'``,
this is obtained from the file (except for ``'RAW'`` files).
channels : int
The number of channels of the file.
If `mode` contains ``'r'``, this is obtained from the file
(except for ``'RAW'`` files).
subtype : str, sometimes optional
The subtype of the sound file. If `mode` contains ``'r'``,
this is obtained from the file (except for ``'RAW'``
files), if not, the default value depends on the selected
`format` (see `default_subtype()`).
See `available_subtypes()` for all possible subtypes for
a given `format`.
endian : {'FILE', 'LITTLE', 'BIG', 'CPU'}, sometimes optional
The endian-ness of the sound file. If `mode` contains
``'r'``, this is obtained from the file (except for
``'RAW'`` files), if not, the default value is ``'FILE'``,
which is correct in most cases.
format : str, sometimes optional
The major format of the sound file. If `mode` contains
``'r'``, this is obtained from the file (except for
``'RAW'`` files), if not, the default value is determined
from the file extension. See `available_formats()` for
all possible values.
closefd : bool, optional
Whether to close the file descriptor on `close()`. Only
applicable if the *file* argument is a file descriptor.
compression_level : float, optional
The compression level on 'write()'. The compression level
should be between 0.0 (minimum compression level) and 1.0
(highest compression level).
See `libsndfile document <https://github.com/libsndfile/libsndfile/blob/c81375f070f3c6764969a738eacded64f53a076e/docs/command.md>`__.
bitrate_mode : {'CONSTANT', 'AVERAGE', 'VARIABLE'}, optional
The bitrate mode on 'write()'.
See `libsndfile document <https://github.com/libsndfile/libsndfile/blob/c81375f070f3c6764969a738eacded64f53a076e/docs/command.md>`__.
Examples
--------
>>> from soundfile import SoundFile
Open an existing file for reading:
>>> myfile = SoundFile('existing_file.wav')
>>> # do something with myfile
>>> myfile.close()
Create a new sound file for reading and writing using a with
statement:
>>> with SoundFile('new_file.wav', 'x+', 44100, 2) as myfile:
>>> # do something with myfile
>>> # ...
>>> assert not myfile.closed
>>> # myfile.close() is called automatically at the end
>>> assert myfile.closed
"""
# resolve PathLike objects (see PEP519 for details):
# can be replaced with _os.fspath(file) for Python >= 3.6
file = file.__fspath__() if hasattr(file, '__fspath__') else file
self._name = file
if mode is None:
mode = getattr(file, 'mode', None)
mode_int = _check_mode(mode)
self._mode = mode
self._compression_level = compression_level
self._bitrate_mode = bitrate_mode
self._info = _create_info_struct(file, mode, samplerate, channels,
format, subtype, endian)
self._file = self._open(file, mode_int, closefd)
if set(mode).issuperset('r+') and self.seekable():
# Move write position to 0 (like in Python file objects)
self.seek(0)
_snd.sf_command(self._file, _snd.SFC_SET_CLIPPING, _ffi.NULL,
_snd.SF_TRUE)
# set compression setting
if self._compression_level is not None:
# needs to be called before set_bitrate_mode
self._set_compression_level(self._compression_level)
if self._bitrate_mode is not None:
self._set_bitrate_mode(self._bitrate_mode)
name = property(lambda self: self._name)
"""The file name of the sound file."""
mode = property(lambda self: self._mode)
"""The open mode the sound file was opened with."""
samplerate = property(lambda self: self._info.samplerate)
"""The sample rate of the sound file."""
frames = property(lambda self: self._info.frames)
"""The number of frames in the sound file."""
channels = property(lambda self: self._info.channels)
"""The number of channels in the sound file."""
format = property(
lambda self: _format_str(self._info.format & _snd.SF_FORMAT_TYPEMASK))
"""The major format of the sound file."""
subtype = property(
lambda self: _format_str(self._info.format & _snd.SF_FORMAT_SUBMASK))
"""The subtype of data in the the sound file."""
endian = property(
lambda self: _format_str(self._info.format & _snd.SF_FORMAT_ENDMASK))
"""The endian-ness of the data in the sound file."""
format_info = property(
lambda self: _format_info(self._info.format &
_snd.SF_FORMAT_TYPEMASK)[1])
"""A description of the major format of the sound file."""
subtype_info = property(
lambda self: _format_info(self._info.format &
_snd.SF_FORMAT_SUBMASK)[1])
"""A description of the subtype of the sound file."""
sections = property(lambda self: self._info.sections)
"""The number of sections of the sound file."""
closed = property(lambda self: self._file is None)
"""Whether the sound file is closed or not."""
_errorcode = property(lambda self: _snd.sf_error(self._file))
"""A pending sndfile error code."""
compression_level = property(lambda self: self._compression_level)
"""The compression level on 'write()'"""
bitrate_mode = property(lambda self: self._bitrate_mode)
"""The bitrate mode on 'write()'"""
@property
def extra_info(self):
"""Retrieve the log string generated when opening the file."""
info = _ffi.new("char[]", 2**14)
_snd.sf_command(self._file, _snd.SFC_GET_LOG_INFO,
info, _ffi.sizeof(info))
return _ffi.string(info).decode('utf-8', 'replace')
# avoid confusion if something goes wrong before assigning self._file:
_file = None
def __repr__(self):
compression_setting = (", compression_level={0}".format(self.compression_level)
if self.compression_level is not None else "")
compression_setting += (", bitrate_mode='{0}'".format(self.bitrate_mode)
if self.bitrate_mode is not None else "")
return ("SoundFile({0.name!r}, mode={0.mode!r}, "
"samplerate={0.samplerate}, channels={0.channels}, "
"format={0.format!r}, subtype={0.subtype!r}, "
"endian={0.endian!r}{1})".format(self, compression_setting))
def __del__(self):
self.close()
def __enter__(self):
return self
def __exit__(self, *args):
self.close()
def __setattr__(self, name, value):
"""Write text meta-data in the sound file through properties."""
if name in _str_types:
self._check_if_closed()
err = _snd.sf_set_string(self._file, _str_types[name],
value.encode())
_error_check(err)
else:
object.__setattr__(self, name, value)
def __getattr__(self, name):
"""Read text meta-data in the sound file through properties."""
if name in _str_types:
self._check_if_closed()
data = _snd.sf_get_string(self._file, _str_types[name])
return _ffi.string(data).decode('utf-8', 'replace') if data else ""
else:
raise AttributeError(
"'SoundFile' object has no attribute {0!r}".format(name))
def __len__(self):
# Note: This is deprecated and will be removed at some point,
# see https://github.com/bastibe/python-soundfile/issues/199
return self._info.frames
def __bool__(self):
# Note: This is temporary until __len__ is removed, afterwards it
# can (and should) be removed without change of behavior
return True
def __nonzero__(self):
# Note: This is only for compatibility with Python 2 and it shall be
# removed at the same time as __bool__().
return self.__bool__()
def seekable(self):
"""Return True if the file supports seeking."""
return self._info.seekable == _snd.SF_TRUE
def seek(self, frames, whence=SEEK_SET):
"""Set the read/write position.
Parameters
----------
frames : int
The frame index or offset to seek.
whence : {SEEK_SET, SEEK_CUR, SEEK_END}, optional
By default (``whence=SEEK_SET``), *frames* are counted from
the beginning of the file.
``whence=SEEK_CUR`` seeks from the current position
(positive and negative values are allowed for *frames*).
``whence=SEEK_END`` seeks from the end (use negative value
for *frames*).
Returns
-------
int
The new absolute read/write position in frames.
Examples
--------
>>> from soundfile import SoundFile, SEEK_END
>>> myfile = SoundFile('stereo_file.wav')
Seek to the beginning of the file:
>>> myfile.seek(0)
0
Seek to the end of the file:
>>> myfile.seek(0, SEEK_END)
44100 # this is the file length
"""
self._check_if_closed()
position = _snd.sf_seek(self._file, frames, whence)
_error_check(self._errorcode)
return position
def tell(self):
"""Return the current read/write position."""
return self.seek(0, SEEK_CUR)
def read(self, frames=-1, dtype='float64', always_2d=False,
fill_value=None, out=None):
"""Read from the file and return data as NumPy array.
Reads the given number of frames in the given data format
starting at the current read/write position. This advances the
read/write position by the same number of frames.
By default, all frames from the current read/write position to
the end of the file are returned.
Use `seek()` to move the current read/write position.
Parameters
----------
frames : int, optional
The number of frames to read. If ``frames < 0``, the whole
rest of the file is read.
dtype : {'float64', 'float32', 'int32', 'int16'}, optional
Data type of the returned array, by default ``'float64'``.
Floating point audio data is typically in the range from
``-1.0`` to ``1.0``. Integer data is in the range from
``-2**15`` to ``2**15-1`` for ``'int16'`` and from
``-2**31`` to ``2**31-1`` for ``'int32'``.
.. note:: Reading int values from a float file will *not*
scale the data to [-1.0, 1.0). If the file contains
``np.array([42.6], dtype='float32')``, you will read
``np.array([43], dtype='int32')`` for
``dtype='int32'``.
Returns
-------
audiodata : `numpy.ndarray` or type(out)
A two-dimensional NumPy (frames x channels) array is
returned. If the sound file has only one channel, a
one-dimensional array is returned. Use ``always_2d=True``
to return a two-dimensional array anyway.
If *out* was specified, it is returned. If *out* has more
frames than available in the file (or if *frames* is
smaller than the length of *out*) and no *fill_value* is
given, then only a part of *out* is overwritten and a view
containing all valid frames is returned.
Other Parameters
----------------
always_2d : bool, optional
By default, reading a mono sound file will return a
one-dimensional array. With ``always_2d=True``, audio data
is always returned as a two-dimensional array, even if the
audio file has only one channel.
fill_value : float, optional
If more frames are requested than available in the file,
the rest of the output is be filled with *fill_value*. If
*fill_value* is not specified, a smaller array is
returned.
out : `numpy.ndarray` or subclass, optional
If *out* is specified, the data is written into the given
array instead of creating a new array. In this case, the
arguments *dtype* and *always_2d* are silently ignored! If
*frames* is not given, it is obtained from the length of
*out*.
Examples
--------
>>> from soundfile import SoundFile
>>> myfile = SoundFile('stereo_file.wav')
Reading 3 frames from a stereo file:
>>> myfile.read(3)
array([[ 0.71329652, 0.06294799],
[-0.26450912, -0.38874483],
[ 0.67398441, -0.11516333]])
>>> myfile.close()
See Also
--------
buffer_read, .write
"""
if out is None:
frames = self._check_frames(frames, fill_value)
out = self._create_empty_array(frames, always_2d, dtype)
else:
if frames < 0 or frames > len(out):
frames = len(out)
frames = self._array_io('read', out, frames)
if len(out) > frames:
if fill_value is None:
out = out[:frames]
else:
out[frames:] = fill_value
return out
def buffer_read(self, frames=-1, dtype=None):
"""Read from the file and return data as buffer object.
Reads the given number of *frames* in the given data format
starting at the current read/write position. This advances the
read/write position by the same number of frames.
By default, all frames from the current read/write position to
the end of the file are returned.
Use `seek()` to move the current read/write position.
Parameters
----------
frames : int, optional
The number of frames to read. If ``frames < 0``, the whole
rest of the file is read.
dtype : {'float64', 'float32', 'int32', 'int16'}
Audio data will be converted to the given data type.
Returns
-------
buffer
A buffer containing the read data.
See Also
--------
buffer_read_into, .read, buffer_write
"""
frames = self._check_frames(frames, fill_value=None)
ctype = self._check_dtype(dtype)
cdata = _ffi.new(ctype + '[]', frames * self.channels)
read_frames = self._cdata_io('read', cdata, ctype, frames)
assert read_frames == frames
return _ffi.buffer(cdata)
def buffer_read_into(self, buffer, dtype):
"""Read from the file into a given buffer object.
Fills the given *buffer* with frames in the given data format
starting at the current read/write position (which can be
changed with `seek()`) until the buffer is full or the end
of the file is reached. This advances the read/write position
by the number of frames that were read.
Parameters
----------
buffer : writable buffer
Audio frames from the file are written to this buffer.
dtype : {'float64', 'float32', 'int32', 'int16'}
The data type of *buffer*.