Image Processing Reference
Matters Concerning Audio
I Hear the Sound of Distant Drums—and Voices
Compressing video deals with the visual part of the movie, but what about the sound
track? This chapter covers the audio standards and formats that are used most often with
compressed video. There are two primary scenarios to consider. There are others, of
course, but these represent the extremes:
Simple conversation with human voices as in telecoms or other low bit-rate
Very-high-quality music, background ambiance, and movie-quality sound tracks.
Before we go any further with our journey to perfect compression, let's divert from
the straight and narrow path and spend a short while listening to the quiet rustle of pix-
els wafting gently in the breeze.
Blowing in the Wind
Audio is a moving pressure wave that the human ear interprets as sound. The change in
pressure is plotted as a waveform and represented electrically as the output of a trans-
ducer. A waveform is a continuously varying value. Although the frequencies are higher,
video is represented in the same way. All the discussion in this chapter regarding sample
rates, word sizes, quantization errors, and digital filtering is relevant for video as well. The
artifacts are the same but appear to be different because we see them instead of hear them.
With video, there are a variety of different sources for the moving picture content, and so
it is with audio. The audio material could have been ingested from any of the following