Source Encoding Problem

Discussion in 'Homework Help' started by Jezzyjez, Jan 4, 2011.

  1. Jezzyjez

    Thread Starter New Member

    Nov 13, 2009
    I am working through some questions and I'm a little unsure on this one:

    Digitisation of a telephone network in the 1960s
    64kbps speech coder
    first filter limit its bandwidth to 3.8kHz and a sampling rate of 8kHz is then applied. The resultant samples are then passed through an A-Law compressor and then quantised to 256 levels (8 bits)

    What would be your main critisism of the scheme?

    My first thought would be that LPF of 3.8kHz may be slightly low and 4kHz would be more suitable. Or further to this a non-uniform quantiser may be used to get better accuracy (but as it is not specified that it is linear can i make this assumption?)

    Any help would be greatly appreciated thanks.