def test_mu_law_encoding(self):
def func(tensor):
qc = 256
return F.mu_law_encoding(tensor, qc)
waveform = common_utils.get_whitenoise()
self._assert_consistency(func, waveform)
def test_contrast(self):
waveform = common_utils.get_whitenoise()
def func(tensor):
enhancement_amount = 80.
return F.contrast(tensor, enhancement_amount)
self._assert_consistency(func, waveform)
def test_dcshift(self):
waveform = common_utils.get_whitenoise()
def func(tensor):
shift = 0.5
limiter_gain = 0.05
return F.dcshift(tensor, shift, limiter_gain)
self._assert_consistency(func, waveform)
def test_overdrive(self):
waveform = common_utils.get_whitenoise()
def func(tensor):
gain = 30.
colour = 50.
return F.overdrive(tensor, gain, colour)
self._assert_consistency(func, waveform)
def test_riaa(self):
if self.dtype == torch.float64:
raise unittest.SkipTest("This test is known to fail for float64")
waveform = common_utils.get_whitenoise(sample_rate=44100)
def func(tensor):
sample_rate = 44100
return F.riaa_biquad(tensor, sample_rate)
self._assert_consistency(func, waveform)
def test_highpass(self):
if self.dtype == torch.float64:
raise unittest.SkipTest("This test is known to fail for float64")
waveform = common_utils.get_whitenoise(sample_rate=44100)
def func(tensor):
sample_rate = 44100
cutoff_freq = 2000.
return F.highpass_biquad(tensor, sample_rate, cutoff_freq)
self._assert_consistency(func, waveform)
def test_bandreject(self):
if self.dtype == torch.float64:
raise unittest.SkipTest("This test is known to fail for float64")
waveform = common_utils.get_whitenoise(sample_rate=44100)
def func(tensor):
sample_rate = 44100
central_freq = 1000.
q = 0.707
return F.bandreject_biquad(tensor, sample_rate, central_freq, q)
self._assert_consistency(func, waveform)
def test_equalizer(self):
if self.dtype == torch.float64:
raise unittest.SkipTest("This test is known to fail for float64")
waveform = common_utils.get_whitenoise(sample_rate=44100)
def func(tensor):
sample_rate = 44100
center_freq = 300.
gain = 1.
q = 0.707
return F.equalizer_biquad(tensor, sample_rate, center_freq, gain,
q)
self._assert_consistency(func, waveform)
def test_spectrogram(self):
def func(tensor):
n_fft = 400
ws = 400
hop = 200
pad = 0
window = torch.hann_window(ws,
device=tensor.device,
dtype=tensor.dtype)
power = 2.
normalize = False
return F.spectrogram(tensor, pad, window, n_fft, hop, ws, power,
normalize)
tensor = common_utils.get_whitenoise()
self._assert_consistency(func, tensor)
def test_perf_biquad_filtering(self):
if self.dtype == torch.float64:
raise unittest.SkipTest("This test is known to fail for float64")
waveform = common_utils.get_whitenoise()
def func(tensor):
a = torch.tensor([0.7, 0.2, 0.6],
device=tensor.device,
dtype=tensor.dtype)
b = torch.tensor([0.4, 0.2, 0.9],
device=tensor.device,
dtype=tensor.dtype)
return F.lfilter(tensor, a, b)
self._assert_consistency(func, waveform)
def test_lfilter(self):
if self.dtype == torch.float64:
raise unittest.SkipTest("This test is known to fail for float64")
waveform = common_utils.get_whitenoise()
def func(tensor):
# Design an IIR lowpass filter using scipy.signal filter design
# https://docs.scipy.org/doc/scipy/reference/generated/scipy.signal.iirdesign.html#scipy.signal.iirdesign
#
# Example
# >>> from scipy.signal import iirdesign
# >>> b, a = iirdesign(0.2, 0.3, 1, 60)
b_coeffs = torch.tensor(
[
0.00299893,
-0.0051152,
0.00841964,
-0.00747802,
0.00841964,
-0.0051152,
0.00299893,
],
device=tensor.device,
dtype=tensor.dtype,
)
a_coeffs = torch.tensor(
[
1.0,
-4.8155751,
10.2217618,
-12.14481273,
8.49018171,
-3.3066882,
0.56088705,
],
device=tensor.device,
dtype=tensor.dtype,
)
return F.lfilter(tensor, a_coeffs, b_coeffs)
self._assert_consistency(func, waveform)
def test_phaser(self):
waveform = common_utils.get_whitenoise(sample_rate=44100)
def func(tensor):
gain_in = 0.5
gain_out = 0.8
delay_ms = 2.0
decay = 0.4
speed = 0.5
sample_rate = 44100
return F.phaser(tensor,
sample_rate,
gain_in,
gain_out,
delay_ms,
decay,
speed,
sinusoidal=True)
self._assert_consistency(func, waveform)
def test_Vol(self):
waveform = common_utils.get_whitenoise()
self._assert_consistency(T.Vol(1.1), waveform)
def test_Fade(self):
waveform = common_utils.get_whitenoise()
fade_in_len = 3000
fade_out_len = 3000
self._assert_consistency(T.Fade(fade_in_len, fade_out_len), waveform)
def test_MuLawEncoding(self):
tensor = common_utils.get_whitenoise()
self._assert_consistency(T.MuLawEncoding(), tensor)
def test_Resample(self):
sr1, sr2 = 16000, 8000
tensor = common_utils.get_whitenoise(sample_rate=sr1)
self._assert_consistency(T.Resample(float(sr1), float(sr2)), tensor)
def test_dither_GPDF(self):
def func(tensor):
return F.dither(tensor, 'GPDF')
tensor = common_utils.get_whitenoise(n_channels=2)
self._assert_consistency(func, tensor, shape_only=True)
