課題を解いた

Author: kuu8902

kkoiso/chapter01/chapter01.py

@@ -0,0 +1,127 @@
+import numpy as np
+import matplotlib.pyplot as plt
+
+#1-1
+a = 1.0
+freq= 440 
+samp_rate = 16000 
+d = 3  
+
+t = np.linspace(0, d, int(samp_rate * d), endpoint=False)
+
+
+sin_wave = a * np.sin(2 * np.pi * freq * t)
+
+
+plt.plot(t,sin_wave)
+plt.title("Sin Wave")
+plt.xlabel("Time [s]")
+plt.ylabel("Amplitude")
+plt.show()
+
+#1-2
+import wave
+
+
+output = 'sin_wave_440Hz.wav'
+with wave.open(output, 'w') as wf:
+    wf.setnchannels(1)  
+    wf.setsampwidth(2)  
+    wf.setframerate(samp_rate)
+    wf.writeframes((sin_wave * 32767).astype(np.int16).tobytes())
+
+
+#1-3
+freq2 = 660  
+sin_wave2 = a * np.sin(2 * np.pi * freq2 * t)
+
+
+stereo_wave = np.vstack((sin_wave, sin_wave2)).T
+
+
+output_stereo = 'stereo_sin_waves.wav'
+with wave.open(output_stereo, 'w') as wf:
+    wf.setnchannels(2) 
+    wf.setsampwidth(2)  
+    wf.setframerate(samp_rate)
+    wf.writeframes((stereo_wave * 32767).astype(np.int16).tobytes())
+
+#1-4
+white_noise = np.random.normal(0, 1, len(t))
+
+
+plt.plot(t, white_noise) 
+plt.title("White Noise")
+plt.xlabel("Time [s]")
+plt.ylabel("Amplitude")
+plt.show()
+
+
+#1-5
+mixed_signal = sin_wave + white_noise
+
+
+plt.plot(t, mixed_signal)  
+plt.title("Mixed Signal (Sin Wave + White Noise)")
+plt.xlabel("Time [s]")
+plt.ylabel("Amplitude")
+plt.show()
+
+#1-6
+def calculate_snr(signal, noise):
+    s_power = np.sum(signal ** 2) / len(signal)
+    n_power = np.sum(noise ** 2) / len(noise)
+    snr = 10 * np.log10(s_power / n_power)
+    return snr
+
+#1-7
+def add_noise_with_snr(signal, desired_snr_db):
+    s_power = np.sum(signal ** 2) / len(signal)
+    snr_linear = 10 ** (desired_snr_db / 10)
+    n_power = s_power / snr_linear
+    noise = np.random.normal(0, np.sqrt(n_power), len(signal))
+    return signal + noise
+
+#1-8
+desired_snr= 6
+noise_signal = add_noise_with_snr(sin_wave, desired_snr)
+
+
+output_noise = 'sin_wave_with_noise_6dB.wav'
+with wave.open(output_noise, 'w') as wf:
+    wf.setnchannels(1)  
+    wf.setsampwidth(2)  
+    wf.setframerate(samp_rate)
+    wf.writeframes((noise_signal * 32767).astype(np.int16).tobytes())
+
+
+#1-9
+from scipy.io import wavfile
+
+rate, data = wavfile.read(output_noise)
+
+
+downsampled_data = data[::2]
+
+
+output_downsampled = 'downsampled_8kHz.wav'
+wavfile.write(output_downsampled, 8000, downsampled_data.astype(np.int16))
+
+#1-10
+filtered_signal = np.convolve(downsampled_data, np.ones(5)/5, mode='valid')
+
+plt.figure(figsize=(14, 6))
+plt.subplot(2, 1, 1)
+plt.plot(downsampled_data[:1000])
+plt.title("Original Downsampled Signal (8 kHz)")
+plt.xlabel("Sample")
+plt.ylabel("Amplitude")
+
+plt.subplot(2, 1, 2)
+plt.plot(filtered_signal[:1000])
+plt.title("Filtered Signal (5-point Moving Average)")
+plt.xlabel("Sample")
+plt.ylabel("Amplitude")
+
+plt.tight_layout()
+plt.show()