def testFloatingPointSampleRateInvalid(self):
with self.cached_session():
with self.assertRaises(TypeError):
ffmpeg.encode_audio(
[[0.0], [1.0]],
file_format='wav',
samples_per_second=12345.678)
def testZeroSampleRateInvalid(self):
with self.cached_session() as sess:
encode_op = ffmpeg.encode_audio(
[[0.0], [1.0]],
file_format='wav',
samples_per_second=0)
with six.assertRaisesRegex(self, Exception, 'must be positive'):
sess.run(encode_op)
def testRoundTripWithPlaceholderSampleRate(self):
with self.cached_session():
placeholder = array_ops.placeholder(dtypes.int32)
audio_op = ffmpeg.decode_audio(
self._contents,
file_format='wav',
samples_per_second=placeholder,
channel_count=1)
encode_op = ffmpeg.encode_audio(
audio_op, file_format='wav', samples_per_second=placeholder)
encoded_contents = encode_op.eval(feed_dict={placeholder: 10000})
self._compareWavFiles(self._contents, encoded_contents)
def testRoundTripWithPlaceholderSampleRate(self):
with self.cached_session():
placeholder = array_ops.placeholder(dtypes.int32)
audio_op = ffmpeg.decode_audio(self._contents,
file_format='wav',
samples_per_second=placeholder,
channel_count=1)
encode_op = ffmpeg.encode_audio(audio_op,
file_format='wav',
samples_per_second=placeholder)
encoded_contents = encode_op.eval(feed_dict={placeholder: 10000})
self._compareWavFiles(self._contents, encoded_contents)
def testRoundTrip(self):
"""Reads a wav file, writes it, and compares them."""
with self.cached_session():
audio_op = ffmpeg.decode_audio(
self._contents,
file_format='wav',
samples_per_second=10000,
channel_count=1)
encode_op = ffmpeg.encode_audio(
audio_op, file_format='wav', samples_per_second=10000)
encoded_contents = encode_op.eval()
self._compareWavFiles(self._contents, encoded_contents)
def testRoundTrip(self):
"""Reads a wav file, writes it, and compares them."""
with self.test_session():
audio_op = ffmpeg.decode_audio(self._contents,
file_format='wav',
samples_per_second=10000,
channel_count=1)
encode_op = ffmpeg.encode_audio(audio_op,
file_format='wav',
samples_per_second=10000)
encoded_contents = encode_op.eval()
self._compareWavFiles(self._contents, encoded_contents)
def save_to_wav(audio_arr_ch1, audio_arr_ch2, sample_rate, original_song_ch1, original_song_ch2, idty, folder, sess):
audio_arr_ch1 = irfft(np.hstack(np.hstack(audio_arr_ch1)))
audio_arr_ch2 = irfft(np.hstack(np.hstack(audio_arr_ch2)))
original_song_ch1 = irfft(np.hstack(np.hstack(original_song_ch1)))
original_song_ch2 = irfft(np.hstack(np.hstack(original_song_ch2)))
original_song = np.hstack(np.array((original_song_ch1, original_song_ch2)).T)
audio_arr = np.hstack(np.array((audio_arr_ch1, audio_arr_ch2)).T)
print(original_song)
w = np.linspace(0, sample_rate, len(audio_arr))
w = w[0:len(audio_arr)]
plt.figure(1)
plt.plot(w, original_song)
plt.savefig(str(folder) + '/original.png')
plt.plot(w, audio_arr)
plt.xlabel('sample')
plt.ylabel('amplitude')
plt.savefig(str(folder) + '/compressed' + str(idty) + '.png')
plt.clf()
cols = 2
rows = math.floor(len(audio_arr)/2)
audio_arr = audio_arr.reshape(rows, cols)
original_song = original_song.reshape(rows, cols)
wav_encoder = ffmpeg.encode_audio(
audio_arr, file_format='wav', samples_per_second=sample_rate)
wav_encoder_orig = ffmpeg.encode_audio(
original_song, file_format='wav', samples_per_second=sample_rate)
wav_file = sess.run(wav_encoder)
wav_orig = sess.run(wav_encoder_orig)
open(str(folder)+'/out.wav', 'wb').write(wav_file)
open(str(folder)+'/wav_orig.wav', 'wb').write(wav_orig)
def testRoundTrip(self):
"""Reads a wav file, writes it, and compares them."""
with self.test_session():
path = os.path.join(
resource_loader.get_data_files_path(), 'testdata/mono_10khz.wav')
with open(path, 'rb') as f:
original_contents = f.read()
audio_op = ffmpeg.decode_audio(
original_contents, file_format='wav', samples_per_second=10000,
channel_count=1)
encode_op = ffmpeg.encode_audio(
audio_op, file_format='wav', samples_per_second=10000)
encoded_contents = encode_op.eval()
self._compareWavFiles(original_contents, encoded_contents)
def testRoundTrip(self):
"""Fabricates some audio, creates a wav file, reverses it, and compares."""
with self.test_session():
path = os.path.join(
resource_loader.get_data_files_path(), 'testdata/mono_10khz.wav')
with open(path, 'r') as f:
original_contents = f.read()
audio_op = ffmpeg.decode_audio(
original_contents, file_format='wav', samples_per_second=10000,
channel_count=1)
encode_op = ffmpeg.encode_audio(
audio_op, file_format='wav', samples_per_second=10000)
encoded_contents = encode_op.eval()
self.assertEqual(original_contents, encoded_contents)
def testRoundTrip(self):
"""Fabricates some audio, creates a wav file, reverses it, and compares."""
with self.test_session():
path = os.path.join(resource_loader.get_data_files_path(),
'testdata/mono_10khz.wav')
with open(path, 'r') as f:
original_contents = f.read()
audio_op = ffmpeg.decode_audio(original_contents,
file_format='wav',
samples_per_second=10000,
channel_count=1)
encode_op = ffmpeg.encode_audio(audio_op,
file_format='wav',
samples_per_second=10000)
encoded_contents = encode_op.eval()
self.assertEqual(original_contents, encoded_contents)
def testFloatingPointSampleRateInvalid(self):
with self.test_session():
with self.assertRaises(TypeError):
ffmpeg.encode_audio([[0.0], [1.0]],
file_format='wav',
samples_per_second=12345.678)
ch2_song[i:i + section_size]
for i in range(0, len(ch2_song), section_size)
]
ch1_song = add_overlap(
ch1_song,
params.overlap_section_size) # [[..], [..], ...] -> [...]
ch2_song = add_overlap(ch2_song, params.overlap_section_size)
# maps sigmoid [0,1] output to [-1,1] for .wav
ch1_song = ((ch1_song * 2) - 1)
ch2_song = ((ch2_song * 2) - 1)
audio_arr = np.hstack(np.array((ch1_song, ch2_song)).T)
cols = 2
rows = math.floor(len(audio_arr) / 2)
audio_arr = audio_arr.reshape(rows, cols)
wav_encoder = ffmpeg.encode_audio(audio_arr,
file_format='wav',
samples_per_second=sample_rate)
wav_file = sess.run(wav_encoder)
with open(f"{params.save_test_output_path_start}{file_number}.wav",
'wb') as f:
f.write(wav_file)
file_number += 1
print('all done')
#https://www.tensorflow.org/api_guides/python/contrib.ffmpeg
import tensorflow as tf
from tensorflow.contrib import ffmpeg
audio_binary = tf.read_file('shibuya.mp3')
waveform = ffmpeg.decode_audio(audio_binary,
file_format='mp3',
samples_per_second=44100,
channel_count=2)
uncompressed_binary = ffmpeg.encode_audio(waveform,
file_format='wav',
samples_per_second=44100)
print(waveform)
print(uncompressed_binary)
