Spaces:
Sleeping
Sleeping
| import math | |
| import os | |
| import random | |
| import paddle | |
| from paddle import nn | |
| import paddle.nn.functional as F | |
| import numpy as np | |
| import librosa | |
| import librosa.util as librosa_util | |
| from librosa.util import normalize, pad_center, tiny | |
| from scipy.signal import get_window | |
| from scipy.io.wavfile import read | |
| 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 paddle.log(paddle.clip(x, min=clip_val) * C) | |
| def dynamic_range_decompression_torch(x, C=1): | |
| """ | |
| PARAMS | |
| ------ | |
| C: compression factor used to compress | |
| """ | |
| return paddle.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 paddle.min(y) < -1.: | |
| print('min value is ', paddle.min(y)) | |
| if paddle.max(y) > 1.: | |
| print('max value is ', paddle.max(y)) | |
| global hann_window | |
| dtype_device = str(y.dtype) + '_' + str(str(y.place)[6:-1]) | |
| wnsize_dtype_device = str(win_size) + '_' + dtype_device | |
| if wnsize_dtype_device not in hann_window: | |
| hann_window[wnsize_dtype_device] = paddle.audio.functional.get_window('hann',win_size).astype(y.dtype) | |
| y = paddle.nn.functional.pad(y.unsqueeze(1), (int((n_fft-hop_size)/2), int((n_fft-hop_size)/2)), mode='reflect', data_format='NCL') | |
| y = y.squeeze(1) | |
| spec = paddle.signal.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) | |
| spec = paddle.as_real(spec) | |
| spec = paddle.sqrt(spec.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.place)[6:-1] | |
| 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] = paddle.to_tensor(mel).astype(spec.dtype) | |
| spec = paddle.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): | |
| if paddle.min(y) < -1.: | |
| print('min value is ', paddle.min(y)) | |
| if paddle.max(y) > 1.: | |
| print('max value is ', paddle.max(y)) | |
| global mel_basis, hann_window | |
| dtype_device = str(y.dtype) + '_' + str(y.place)[6:-1] | |
| 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] = paddle.to_tensor(mel).astype(y.dtype) | |
| if wnsize_dtype_device not in hann_window: | |
| hann_window[wnsize_dtype_device] = paddle.audio.functional.get_window('hann',win_size).astype(y.dtype) | |
| y = paddle.nn.functional.pad(y.unsqueeze(1), (int((n_fft-hop_size)/2), int((n_fft-hop_size)/2)), mode='reflect',data_format = 'NCL') | |
| y = y.squeeze(1) | |
| spec = paddle.signal.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) | |
| spec = paddle.as_real(spec) | |
| spec = paddle.sqrt(spec.pow(2).sum(-1) + 1e-6) | |
| spec = paddle.matmul(mel_basis[fmax_dtype_device], spec) | |
| spec = spectral_normalize_torch(spec) | |
| return spec | |