====================================
 Real Time Processing Sound Signals
====================================

First we add a plot of the powerspectrum of the FFT of the sound
signal. In the figure below a screenshot is shown of Miles Davis. The
typical pattern of harmonics (frequency spikes at constant interval)
of a wind instrument are shown (remember the clarinet exercise).

.. figure:: /figures/miles_so_what_visualization.png

   **Screenshot of the real time audio visualization.**


.. literalinclude:: read_play_fft.py
   :language: python


Observe that for each processed block of data an FFT is
calculated. Quite some work to be done by the processor.


In the second example we implement a digital low pass filter to
process the stream of audio data. Both left and right channel are
filtered in case of a stereo signal. Observe that the 'high hat' on
the drums is barely heard now.



.. figure:: /figures/miles_so_what_lowpass.png

   **Screenshot of the real time audio visualization** of a low pass
   filtered signal. Observe that in comparison with the unfiltered
   signal we have lost some of the higher frequency harmonics of
   Miles' trumpet.

.. literalinclude:: play_process_sound.py
   :language: python