Spaces:
Running
Running
| import os | |
| import numpy as np | |
| from scipy.io import wavfile | |
| from scipy import signal | |
| import resampy | |
| from hparams import hparams as hp | |
| def load_wav(path, sr): | |
| orig_sr, audio = wavfile.read(path) | |
| if len(audio) < 100: # Arbitrary threshold (can be higher for safety) | |
| raise ValueError(f"Input audio too short: {len(audio)} samples") | |
| if audio.dtype.kind == 'i': | |
| audio = audio.astype(np.float32) / np.iinfo(audio.dtype).max | |
| else: | |
| audio = audio.astype(np.float32) | |
| if orig_sr != sr: | |
| audio = resampy.resample(audio, orig_sr, sr) | |
| return audio | |
| def save_wav(wav, path, sr): | |
| """ | |
| Save a float32 waveform to disk as 16-bit PCM WAV. | |
| """ | |
| wav_int16 = (wav * 32767).clip(-32767, 32767).astype(np.int16) | |
| wavfile.write(path, sr, wav_int16) | |
| def preemphasis(wav, k, preemphasize=True): | |
| if preemphasize: | |
| return signal.lfilter([1, -k], [1], wav) | |
| return wav | |
| def inv_preemphasis(wav, k, inv_preemphasize=True): | |
| if inv_preemphasize: | |
| return signal.lfilter([1], [1, -k], wav) | |
| return wav | |
| def get_hop_size(): | |
| hop_size = hp.hop_size | |
| if hop_size is None: | |
| assert hp.frame_shift_ms is not None | |
| hop_size = int(hp.frame_shift_ms / 1000 * hp.sample_rate) | |
| return hop_size | |
| def linearspectrogram(wav): | |
| D = _stft(preemphasis(wav, hp.preemphasis, hp.preemphasize)) | |
| S = _amp_to_db(np.abs(D)) - hp.ref_level_db | |
| return _normalize(S) if hp.signal_normalization else S | |
| def melspectrogram(wav): | |
| D = _stft(preemphasis(wav, hp.preemphasis, hp.preemphasize)) | |
| S = _amp_to_db(_linear_to_mel(np.abs(D))) - hp.ref_level_db | |
| return _normalize(S) if hp.signal_normalization else S | |
| def _lws_processor(): | |
| import lws | |
| return lws.lws(hp.n_fft, get_hop_size(), fftsize=hp.win_size, mode="speech") | |
| def _stft(y): | |
| if hp.use_lws: | |
| return _lws_processor().stft(y).T | |
| else: | |
| import librosa # Safe to import inside function | |
| return librosa.stft(y=y, n_fft=hp.n_fft, hop_length=get_hop_size(), win_length=hp.win_size) | |
| def num_frames(length, fsize, fshift): | |
| pad = (fsize - fshift) | |
| if length % fshift == 0: | |
| M = (length + pad * 2 - fsize) // fshift + 1 | |
| else: | |
| M = (length + pad * 2 - fsize) // fshift + 2 | |
| return M | |
| def pad_lr(x, fsize, fshift): | |
| M = num_frames(len(x), fsize, fshift) | |
| pad = (fsize - fshift) | |
| T = len(x) + 2 * pad | |
| r = (M - 1) * fshift + fsize - T | |
| return pad, pad + r | |
| def librosa_pad_lr(x, fsize, fshift): | |
| return 0, (x.shape[0] // fshift + 1) * fshift - x.shape[0] | |
| _mel_basis = None | |
| def _linear_to_mel(spectrogram): | |
| global _mel_basis | |
| if _mel_basis is None: | |
| _mel_basis = _build_mel_basis() | |
| return np.dot(_mel_basis, spectrogram) | |
| def _build_mel_basis(): | |
| import librosa.filters # Imported only when needed | |
| assert hp.fmax <= hp.sample_rate // 2 | |
| return librosa.filters.mel( | |
| sr=hp.sample_rate, | |
| n_fft=hp.n_fft, | |
| n_mels=hp.num_mels, | |
| fmin=hp.fmin, | |
| fmax=hp.fmax | |
| ) | |
| def _amp_to_db(x): | |
| min_level = np.exp(hp.min_level_db / 20 * np.log(10)) | |
| return 20 * np.log10(np.maximum(min_level, x)) | |
| def _db_to_amp(x): | |
| return np.power(10.0, x * 0.05) | |
| def _normalize(S): | |
| if hp.allow_clipping_in_normalization: | |
| if hp.symmetric_mels: | |
| return np.clip((2 * hp.max_abs_value) * ((S - hp.min_level_db) / (-hp.min_level_db)) - hp.max_abs_value, | |
| -hp.max_abs_value, hp.max_abs_value) | |
| else: | |
| return np.clip(hp.max_abs_value * ((S - hp.min_level_db) / (-hp.min_level_db)), 0, hp.max_abs_value) | |
| assert S.max() <= 0 and S.min() - hp.min_level_db >= 0 | |
| if hp.symmetric_mels: | |
| return (2 * hp.max_abs_value) * ((S - hp.min_level_db) / (-hp.min_level_db)) - hp.max_abs_value | |
| else: | |
| return hp.max_abs_value * ((S - hp.min_level_db) / (-hp.min_level_db)) | |
| def _denormalize(D): | |
| if hp.allow_clipping_in_normalization: | |
| if hp.symmetric_mels: | |
| return (((np.clip(D, -hp.max_abs_value, | |
| hp.max_abs_value) + hp.max_abs_value) * -hp.min_level_db / (2 * hp.max_abs_value)) | |
| + hp.min_level_db) | |
| else: | |
| return ((np.clip(D, 0, hp.max_abs_value) * -hp.min_level_db / hp.max_abs_value) + hp.min_level_db) | |
| if hp.symmetric_mels: | |
| return (((D + hp.max_abs_value) * -hp.min_level_db / (2 * hp.max_abs_value)) + hp.min_level_db) | |
| else: | |
| return ((D * -hp.min_level_db / hp.max_abs_value) + hp.min_level_db) | |