Spaces:
Running
Running
File size: 7,954 Bytes
9ace58a |
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 |
"""
Module to read / write wav files using numpy arrays
Functions
---------
`read`: Return the sample rate (in samples/sec) and data from a WAV file.
`write`: Write a numpy array as a WAV file.
"""
from __future__ import division, print_function, absolute_import
import sys
import numpy
import struct
import warnings
import os
class WavFileWarning(UserWarning):
pass
_big_endian = False
WAVE_FORMAT_PCM = 0x0001
WAVE_FORMAT_IEEE_FLOAT = 0x0003
WAVE_FORMAT_EXTENSIBLE = 0xfffe
KNOWN_WAVE_FORMATS = (WAVE_FORMAT_PCM, WAVE_FORMAT_IEEE_FLOAT)
# assumes file pointer is immediately
# after the 'fmt ' id
def _read_fmt_chunk(fid):
if _big_endian:
fmt = '>'
else:
fmt = '<'
res = struct.unpack(fmt+'iHHIIHH',fid.read(20))
size, comp, noc, rate, sbytes, ba, bits = res
if comp not in KNOWN_WAVE_FORMATS or size > 16:
comp = WAVE_FORMAT_PCM
warnings.warn("Unknown wave file format", WavFileWarning)
if size > 16:
fid.read(size - 16)
return size, comp, noc, rate, sbytes, ba, bits
# assumes file pointer is immediately
# after the 'data' id
def _read_data_chunk(fid, comp, noc, bits, mmap=False):
if _big_endian:
fmt = '>i'
else:
fmt = '<i'
size = struct.unpack(fmt,fid.read(4))[0]
bytes = bits//8
if bits == 8:
dtype = 'u1'
else:
if _big_endian:
dtype = '>'
else:
dtype = '<'
if comp == 1:
dtype += 'i%d' % bytes
else:
dtype += 'f%d' % bytes
if not mmap:
data = numpy.fromstring(fid.read(size), dtype=dtype)
else:
start = fid.tell()
data = numpy.memmap(fid, dtype=dtype, mode='c', offset=start,
shape=(size//bytes,))
fid.seek(start + size)
if noc > 1:
data = data.reshape(-1,noc)
return data
def _skip_unknown_chunk(fid):
if _big_endian:
fmt = '>i'
else:
fmt = '<i'
data = fid.read(4)
size = struct.unpack(fmt, data)[0]
fid.seek(size, 1)
def _read_riff_chunk(fid):
global _big_endian
str1 = fid.read(4)
if str1 == b'RIFX':
_big_endian = True
elif str1 != b'RIFF':
raise ValueError("Not a WAV file.")
if _big_endian:
fmt = '>I'
else:
fmt = '<I'
fsize = struct.unpack(fmt, fid.read(4))[0] + 8
str2 = fid.read(4)
if (str2 != b'WAVE'):
raise ValueError("Not a WAV file.")
if str1 == b'RIFX':
_big_endian = True
return fsize
# open a wave-file
def read(filename, mmap=False):
"""
Return the sample rate (in samples/sec) and data from a WAV file
Parameters
----------
filename : string or open file handle
Input wav file.
mmap : bool, optional
Whether to read data as memory mapped.
Only to be used on real files (Default: False)
.. versionadded:: 0.12.0
Returns
-------
rate : int
Sample rate of wav file
data : numpy array
Data read from wav file
Notes
-----
* The file can be an open file or a filename.
* The returned sample rate is a Python integer
* The data is returned as a numpy array with a
data-type determined from the file.
"""
if hasattr(filename,'read'):
fid = filename
mmap = False
else:
fid = open(filename, 'rb')
try:
# some files seem to have the size recorded in the header greater than
# the actual file size.
fid.seek(0, os.SEEK_END)
actual_size = fid.tell()
fid.seek(0)
fsize = _read_riff_chunk(fid)
# the fsize should be identical to the actual size, if not
# the header information is wrong and we need to correct it.
if fsize != actual_size:
fsize = actual_size
noc = 1
bits = 8
comp = WAVE_FORMAT_PCM
while (fid.tell() < fsize):
# read the next chunk
chunk_id = fid.read(4)
if chunk_id == b'fmt ':
size, comp, noc, rate, sbytes, ba, bits = _read_fmt_chunk(fid)
elif chunk_id == b'fact':
_skip_unknown_chunk(fid)
elif chunk_id == b'data':
data = _read_data_chunk(fid, comp, noc, bits, mmap=mmap)
elif chunk_id == b'LIST':
# Someday this could be handled properly but for now skip it
_skip_unknown_chunk(fid)
# OMA warning - I've commented out the following lines
# else:
# warnings.warn("Chunk (non-data) not understood, skipping it.", WavFileWarning)
# _skip_unknown_chunk(fid)
finally:
if not hasattr(filename,'read'):
fid.close()
else:
fid.seek(0)
return rate, data
# Write a wave-file
# sample rate, data
def write(filename, rate, data):
"""
Write a numpy array as a WAV file
Parameters
----------
filename : string or open file handle
Output wav file
rate : int
The sample rate (in samples/sec).
data : ndarray
A 1-D or 2-D numpy array of either integer or float data-type.
Notes
-----
* The file can be an open file or a filename.
* Writes a simple uncompressed WAV file.
* The bits-per-sample will be determined by the data-type.
* To write multiple-channels, use a 2-D array of shape
(Nsamples, Nchannels).
"""
if hasattr(filename, 'write'):
fid = filename
else:
fid = open(filename, 'wb')
try:
# kind of numeric data in the numpy array
dkind = data.dtype.kind
if not (dkind == 'i' or dkind == 'f' or (dkind == 'u' and data.dtype.itemsize == 1)):
raise ValueError("Unsupported data type '%s'" % data.dtype)
# wav header stuff
# http://soundfile.sapp.org/doc/WaveFormat/
fid.write(b'RIFF')
# placeholder for chunk size (updated later)
fid.write(b'\x00\x00\x00\x00')
fid.write(b'WAVE')
# fmt chunk
fid.write(b'fmt ')
if dkind == 'f':
# comp stands for compression. PCM = 1
comp = 3
else:
comp = 1
# determine number of channels
if data.ndim == 1:
noc = 1
else:
noc = data.shape[1]
bits = data.dtype.itemsize * 8
# number of bytes per second, at the specified sampling rate rate,
# bits per sample and number of channels (just needed for wav header)
sbytes = rate*(bits // 8)*noc
# number of bytes per sample
ba = noc * (bits // 8)
# https://docs.python.org/3/library/struct.html#struct-format-strings
# Write the data (16, comp, noc, etc) in the correct binary format
# for the wav header. the string format (first arg) specifies how many bytes for each
# value.
fid.write(struct.pack('<ihHIIHH', 16, comp, noc, rate, sbytes, ba, bits))
# data chunk: the word 'data' followed by the size followed by the actual data
fid.write(b'data')
fid.write(struct.pack('<i', data.nbytes))
if data.dtype.byteorder == '>' or (data.dtype.byteorder == '=' and sys.byteorder == 'big'):
data = data.byteswap()
_array_tofile(fid, data)
# Determine file size and place it in correct
# position at start of the file (replacing the 4 bytes of zeros)
size = fid.tell()
fid.seek(4)
fid.write(struct.pack('<i', size-8))
finally:
if not hasattr(filename,'write'):
fid.close()
else:
fid.seek(0)
if sys.version_info[0] >= 3:
def _array_tofile(fid, data):
# ravel gives a c-contiguous buffer
fid.write(data.ravel().view('b').data)
else:
def _array_tofile(fid, data):
fid.write(data.tostring())
|