add comments back

system/micd.py

@@ -13,6 +13,8 @@
 SAMPLE_BUFFER = 4096 # (approx 100ms)
 
 def precompute_a_weighting():
+  # Calculate the A-weighting filter
+  # https://en.wikipedia.org/wiki/A-weighting
   freqs = np.fft.fftfreq(FFT_SAMPLES, d=1 / SAMPLE_RATE)
   A = 12194 ** 2 * freqs ** 4 / ((freqs ** 2 + 20.6 ** 2) * (freqs ** 2 + 12194 ** 2) * np.sqrt((freqs ** 2 + 107.7 ** 2) * (freqs ** 2 + 737.9 ** 2)))
   return A / np.max(A)
@@ -33,6 +35,8 @@ def calculate_spl(measurements):
 def apply_a_weighting(measurements: np.ndarray) -> np.ndarray:
   # Generate a Hanning window of the same length as the audio measurements
   measurements_windowed = measurements * np.hanning(len(measurements))
+
+  # Apply the A-weighting filter to the signal
   weighted = np.abs(np.fft.ifft(np.fft.fft(measurements_windowed) * A_WEIGHTING))
   return weighted