def convolve_and_pad(wav, h_list):
data_list = []
for i_h, h in enumerate(h_list):
y = au.convolve(wav, h)
y_length = au.next_power_of_two(np.size(y))
data = au.pad_to(y, y_length, 0)
data_list.append(data)
return np.array(data_list)
def generate_waveforms(wav, h_list):
data_list = []
target_list = []
for i_h, h in enumerate(h_list):
y = au.convolve(wav, h)
y_length = au.next_power_of_two(np.size(y))
data = au.pad_to(y, y_length, 0)
target = au.pad_to(h, y_length, 0)
target_list.append(target)
data_list.append(data)
return np.array(target_list), np.array(data_list)
def pad_list_to_pow2(h_list):
longest_irf = len(max(h_list, key=len))
target_length = au.next_power_of_two(longest_irf)
h_list = [au.pad_to(h, target_length) for h in h_list]
return h_list
import rirnet.acoustic_utils as au
import librosa
import numpy as np
import matplotlib.pyplot as plt
out_path = '../audio/chamber/'
in_path = '../audio/chamber/full/full.wav'
rate = 44100
data, rate = au.read_wav(in_path, rate=rate)
sound_starts = librosa.onset.onset_detect(data, sr=rate, backtrack=True) * 512
for i, start in enumerate(sound_starts):
stop = start + au.next_power_of_two(int(rate / 4))
energy = np.sum(np.abs(data[stop - 100:stop]))
if energy < 0.01:
au.save_wav(out_path + 'ch_{}.wav'.format(i), data[start:stop], rate)