Use torchaudio melscale 'slaney' instead of librosa in WaveRNN pipeline preprocessing

[discort]

cc #593

[discort]

cc @mthrok @vincentqb

[mthrok]

If you have an extra bandwidth, can you add a test that verifies that librosa.feature.melspectrogram(S=spectrogram, ...) and MelScale()(spectrogram) produce the same result?

We do have a test for MelSpectrogram here

audio/test/torchaudio_unittest/transforms/librosa_compatibility_test_impl.py

Lines 62 to 87 in 35d68fd

	@nested_params(
	[
	param(n_fft=400, hop_length=200, n_mels=64),
	param(n_fft=600, hop_length=100, n_mels=128),
	param(n_fft=200, hop_length=50, n_mels=32),
	],
	[param(norm=norm) for norm in [None, 'slaney']],
	[param(mel_scale=mel_scale) for mel_scale in ['htk', 'slaney']],
	)
	def test_MelSpectrogram(self, n_fft, hop_length, n_mels, norm, mel_scale):
	sample_rate = 16000
	waveform = get_sinusoid(
	sample_rate=sample_rate, n_channels=1,
	).to(self.device, self.dtype)
	expected = librosa.feature.melspectrogram(
	y=waveform[0].cpu().numpy(),
	sr=sample_rate, n_fft=n_fft,
	hop_length=hop_length, n_mels=n_mels, norm=norm,
	htk=mel_scale == "htk")
	result = T.MelSpectrogram(
	sample_rate=sample_rate, window_fn=torch.hann_window,
	hop_length=hop_length, n_mels=n_mels,
	n_fft=n_fft, norm=norm, mel_scale=mel_scale,
	).to(self.device, self.dtype)(waveform)[0]
	self.assertEqual(result, torch.from_numpy(expected), atol=5e-4, rtol=1e-5)

It will be nice to add one for MelScale as well.

[mthrok]

Talking about MelSpectrogram, how about using MelSpectrogram directly?

[discort]

Talking about MelSpectrogram, how about using MelSpectrogram directly?

Makes sense.
Fixed it and checked manually.

[mthrok]

@vincentqb Review please.

[vincentqb]

LGTM, but we need to make sure the convergence of the model has not been affected. @discort -- have you reran the training loop to see? I'm rerunning the model on my side to see if the convergence profile changed.

[discort]

thanks for getting back to me @vincentqb
Unfortunately I don't have resources to check the convergency. Let me know if there are any problems and next week I'll try to allocate some staff to check it manually.

[vincentqb]

Alright, LGTM, thanks!

[vincentqb]

Thanks!