audio.py

import librosa
import librosa.filters
import math
import numpy as np
import scipy
import hparams


# Find end point need trans spec
# inv spec needn't trans spec


def load_wav(path):
    return librosa.core.load(path, sr=hparams.sample_rate)[0]


def save_wav(wav, path):
    wav *= 32767 / max(0.01, np.max(np.abs(wav)))
    scipy.io.wavfile.write(path, hparams.sample_rate, wav.astype(np.int16))


def preemphasis(x):
    return scipy.signal.lfilter([1, -hparams.preemphasis], [1], x)


def inv_preemphasis(x):
    return scipy.signal.lfilter([1], [1, -hparams.preemphasis], x)


def spectrogram(y):
    D = _stft(preemphasis(y))
    S = _amp_to_db(np.abs(D)) - hparams.ref_level_db
    return _normalize(S)


def inv_spectrogram(spectrogram):
    '''Converts spectrogram to waveform using librosa'''
    S = _db_to_amp(_denormalize(spectrogram) +
                   hparams.ref_level_db)  # Convert back to linear
    # Reconstruct phase
    return inv_preemphasis(_griffin_lim(S ** hparams.power))


def melspectrogram(y):
    D = _stft(preemphasis(y))
    S = _amp_to_db(_linear_to_mel(np.abs(D))) - hparams.ref_level_db

    return _normalize(S)


def find_endpoint(wav, threshold_db=-40, min_silence_sec=0.8):
    window_length = int(hparams.sample_rate * min_silence_sec)
    hop_length = int(window_length / 4)
    threshold = _db_to_amp(threshold_db)
    for x in range(hop_length, len(wav) - window_length, hop_length):
        if np.max(wav[x:x+window_length]) < threshold:
            return x + hop_length
    return len(wav)


def _griffin_lim(S):
    '''librosa implementation of Griffin-Lim
    Based on https://github.com/librosa/librosa/issues/434
    '''
    angles = np.exp(2j * np.pi * np.random.rand(*S.shape))
    S_complex = np.abs(S).astype(np.complex)
    y = _istft(S_complex * angles)
    for i in range(hparams.griffin_lim_iters):
        angles = np.exp(1j * np.angle(_stft(y)))
        y = _istft(S_complex * angles)
    return y


def _stft(y):
    n_fft, hop_length, win_length = _stft_parameters()
    return librosa.stft(y=y, n_fft=n_fft, hop_length=hop_length, win_length=win_length)


def _istft(y):
    _, hop_length, win_length = _stft_parameters()
    return librosa.istft(y, hop_length=hop_length, win_length=win_length)


def _stft_parameters():
    n_fft = (hparams.num_freq - 1) * 2

    # hop_length = int(hparams.frame_shift_ms / 1000 * hparams.sample_rate)
    # win_length = int(hparams.frame_length_ms / 1000 * hparams.sample_rate)

    hop_length = hparams.hop_length
    win_length = hparams.win_length

    # print(hop_length, win_length)

    return n_fft, hop_length, win_length


# Conversions:
_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():
    n_fft = (hparams.num_freq - 1) * 2
    return librosa.filters.mel(hparams.sample_rate, n_fft, n_mels=hparams.num_mels)


def _amp_to_db(x):
    return 20 * np.log10(np.maximum(1e-5, x))


def _db_to_amp(x):
    return np.power(10.0, x * 0.05)


def _normalize(S):
    return np.clip((S - hparams.min_level_db) / -hparams.min_level_db, 0, 1)


def _denormalize(S):
    return (np.clip(S, 0, 1) * -hparams.min_level_db) + hparams.min_level_db


# def get_hop_size():
#     hop_size = int(hparams.frame_shift_ms / 1000 * hparams.sample_rate)
#     return hop_size


# def get_win_size():
#     win_size = int(hparams.frame_length_ms / 1000 * hparams.sample_rate)
#     return win_size


_inv_mel_basis = None


def _mel_to_linear(mel_spectrogram):
    global _inv_mel_basis
    if _inv_mel_basis is None:
        _inv_mel_basis = np.linalg.pinv(_build_mel_basis())
    return np.maximum(1e-10, np.dot(_inv_mel_basis, mel_spectrogram))


def inv_mel_spectrogram(mel_spectrogram):
    '''Converts mel spectrogram to waveform using librosa'''
    if hparams.signal_normalization:
        D = _denormalize(mel_spectrogram)
    else:
        D = mel_spectrogram

    # Convert back to linear
    S = _mel_to_linear(_db_to_amp(D + hparams.ref_level_db))

    return inv_preemphasis(_griffin_lim(S ** hparams.power))