Telephone Dialing System

import numpy as np import matplotlib.pyplot as plt from scipy.io import wavfile from scipy import signal from scipy.fft import rfft, rfftfreq, irfft import IPython from pathlib import Path %matplotlib inline plt.rcParams['figure.figsize'] = (16.0, 9.0) # set default size of plots plt.rcParams['image.interpolation'] = 'nearest' plt.rcParams['image.cmap'] = 'gray'

def plot(time, signal, save=None, xlabel='', ylabel=''): plt.plot(time, signal) # plt.legend() if ylabel: plt.ylabel(ylabel) if xlabel: plt.xlabel(xlabel) if save: Path("imgs/").mkdir(exist_ok=True) plt.savefig('imgs/' + save, bbox_inches='tight') plt.show()

freq = [ (697, 1209), # 1 (770, 1477), # 6 (852, 1336), # 8 ] SOUND_D = 0.15 SILENCE_D = 0.05 x = get_time_linespace(len(freq)*(SOUND_D + SILENCE_D)+SILENCE_D) y = np.zeros_like(get_time_linespace(SILENCE_D)) for f1, f2 in freq: y = np.append(y, np.sin(2*np.pi*f1*get_time_linespace(SOUND_D)) + \ np.sin(2*np.pi*f2*get_time_linespace(SOUND_D))) y = np.append(y, np.zeros_like(get_time_linespace(SILENCE_D))) y /= np.max(np.abs(y))

print(x.shape, y.shape)

plot(x, y, save='sound_sig.pdf', xlabel='Time [sec]', ylabel='Amplitude')

wavfile.write('new_data.wav', samplerate, y.astype(np.float32)) IPython.display.Audio("new_data.wav")

time_slice = np.where((0.10 <= x) & (x <= 0.12)) plot(x[time_slice], y[time_slice], 'magnified.pdf')

N = x.shape[0] yf = rfft(y) * 2 / N xf = rfftfreq(N, 1 / samplerate) freq_slice = np.where(xf <= 2e3) plot(xf[freq_slice], np.abs(yf[freq_slice]), 'ftt.pdf', xlabel='Frequency [Hz]', ylabel='Amplitude')

f, t, Sxx = signal.spectrogram(y, samplerate, window=np.ones(2048)) freq_slice = np.where(f <= 2e3) f, Sxx = f[freq_slice], Sxx[freq_slice,:][0] plt.pcolormesh(t, f, Sxx, shading='gouraud', cmap='jet') plt.ylabel('Frequency [Hz]') plt.xlabel('Time [sec]') plt.colorbar() plt.savefig('imgs/spec.pdf', bbox_inches='tight') plt.show()