File size: 5,868 Bytes
4300fed
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
import torch
import torch.utils.data
import librosa
from librosa.filters import mel as librosa_mel_fn

MAX_WAV_VALUE = 32768.0


def dynamic_range_compression_torch(x, C=1, clip_val=1e-5):
    """
    PARAMS
    ------
    C: compression factor
    """
    return torch.log(torch.clamp(x, min=clip_val) * C)


def dynamic_range_decompression_torch(x, C=1):
    """
    PARAMS
    ------
    C: compression factor used to compress
    """
    return torch.exp(x) / C


def spectral_normalize_torch(magnitudes):
    output = dynamic_range_compression_torch(magnitudes)
    return output


def spectral_de_normalize_torch(magnitudes):
    output = dynamic_range_decompression_torch(magnitudes)
    return output


mel_basis = {}
hann_window = {}


def spectrogram_torch(y, n_fft, sampling_rate, hop_size, win_size, center=False):
    if torch.min(y) < -1.1:
        print("min value is ", torch.min(y))
    if torch.max(y) > 1.1:
        print("max value is ", torch.max(y))

    global hann_window
    dtype_device = str(y.dtype) + "_" + str(y.device)
    wnsize_dtype_device = str(win_size) + "_" + dtype_device
    if wnsize_dtype_device not in hann_window:
        hann_window[wnsize_dtype_device] = torch.hann_window(win_size).to(
            dtype=y.dtype, device=y.device
        )

    y = torch.nn.functional.pad(
        y.unsqueeze(1),
        (int((n_fft - hop_size) / 2), int((n_fft - hop_size) / 2)),
        mode="reflect",
    )
    y = y.squeeze(1)

    spec = torch.stft(
        y,
        n_fft,
        hop_length=hop_size,
        win_length=win_size,
        window=hann_window[wnsize_dtype_device],
        center=center,
        pad_mode="reflect",
        normalized=False,
        onesided=True,
        return_complex=False,
    )

    spec = torch.sqrt(spec.pow(2).sum(-1) + 1e-6)
    return spec


def spectrogram_torch_conv(y, n_fft, sampling_rate, hop_size, win_size, center=False):
    global hann_window
    dtype_device = str(y.dtype) + '_' + str(y.device)
    wnsize_dtype_device = str(win_size) + '_' + dtype_device
    if wnsize_dtype_device not in hann_window:
        hann_window[wnsize_dtype_device] = torch.hann_window(win_size).to(dtype=y.dtype, device=y.device)

    y = torch.nn.functional.pad(y.unsqueeze(1), (int((n_fft-hop_size)/2), int((n_fft-hop_size)/2)), mode='reflect')
    
    # ******************** original ************************#
    # y = y.squeeze(1)
    # spec1 = torch.stft(y, n_fft, hop_length=hop_size, win_length=win_size, window=hann_window[wnsize_dtype_device],
    #                   center=center, pad_mode='reflect', normalized=False, onesided=True, return_complex=False)

    # ******************** ConvSTFT ************************#
    freq_cutoff = n_fft // 2 + 1
    fourier_basis = torch.view_as_real(torch.fft.fft(torch.eye(n_fft)))
    forward_basis = fourier_basis[:freq_cutoff].permute(2, 0, 1).reshape(-1, 1, fourier_basis.shape[1])
    forward_basis = forward_basis * torch.as_tensor(librosa.util.pad_center(torch.hann_window(win_size), size=n_fft)).float()

    import torch.nn.functional as F

    # if center:
    #     signal = F.pad(y[:, None, None, :], (n_fft // 2, n_fft // 2, 0, 0), mode = 'reflect').squeeze(1)
    assert center is False

    forward_transform_squared = F.conv1d(y, forward_basis.to(y.device), stride = hop_size)
    spec2 = torch.stack([forward_transform_squared[:, :freq_cutoff, :], forward_transform_squared[:, freq_cutoff:, :]], dim = -1)


    # ******************** Verification ************************#
    spec1 = torch.stft(y.squeeze(1), n_fft, hop_length=hop_size, win_length=win_size, window=hann_window[wnsize_dtype_device],
                      center=center, pad_mode='reflect', normalized=False, onesided=True, return_complex=False)
    assert torch.allclose(spec1, spec2, atol=1e-4)

    spec = torch.sqrt(spec2.pow(2).sum(-1) + 1e-6)
    return spec


def spec_to_mel_torch(spec, n_fft, num_mels, sampling_rate, fmin, fmax):
    global mel_basis
    dtype_device = str(spec.dtype) + "_" + str(spec.device)
    fmax_dtype_device = str(fmax) + "_" + dtype_device
    if fmax_dtype_device not in mel_basis:
        mel = librosa_mel_fn(sr=sampling_rate, n_fft=n_fft, n_mels=num_mels, fmin=fmin, fmax=fmax)
        mel_basis[fmax_dtype_device] = torch.from_numpy(mel).to(
            dtype=spec.dtype, device=spec.device
        )
    spec = torch.matmul(mel_basis[fmax_dtype_device], spec)
    spec = spectral_normalize_torch(spec)
    return spec


def mel_spectrogram_torch(
    y, n_fft, num_mels, sampling_rate, hop_size, win_size, fmin, fmax, center=False
):
    global mel_basis, hann_window
    dtype_device = str(y.dtype) + "_" + str(y.device)
    fmax_dtype_device = str(fmax) + "_" + dtype_device
    wnsize_dtype_device = str(win_size) + "_" + dtype_device
    if fmax_dtype_device not in mel_basis:
        mel = librosa_mel_fn(sr=sampling_rate, n_fft=n_fft, n_mels=num_mels, fmin=fmin, fmax=fmax)
        mel_basis[fmax_dtype_device] = torch.from_numpy(mel).to(
            dtype=y.dtype, device=y.device
        )
    if wnsize_dtype_device not in hann_window:
        hann_window[wnsize_dtype_device] = torch.hann_window(win_size).to(
            dtype=y.dtype, device=y.device
        )

    y = torch.nn.functional.pad(
        y.unsqueeze(1),
        (int((n_fft - hop_size) / 2), int((n_fft - hop_size) / 2)),
        mode="reflect",
    )
    y = y.squeeze(1)

    spec = torch.stft(
        y,
        n_fft,
        hop_length=hop_size,
        win_length=win_size,
        window=hann_window[wnsize_dtype_device],
        center=center,
        pad_mode="reflect",
        normalized=False,
        onesided=True,
        return_complex=False,
    )

    spec = torch.sqrt(spec.pow(2).sum(-1) + 1e-6)

    spec = torch.matmul(mel_basis[fmax_dtype_device], spec)
    spec = spectral_normalize_torch(spec)

    return spec