On Nov 16, 11:12*pm, jim wrote:
The bandwidth of any communication channel - whether formed by CAT5, 6
or fiberoptic or radio waves or plain audio or whatever - is a
physical limit of that channel (often expressed in MHz). *No amount of
jiggery pokery with the electronics, compression algorithm maths,
frequency mixing, or any other deceptively clever compression scams,
can increase the maximum physical bandwidth.

ISTM you are confusing *bit rate* with *symbol rate*. Yes, the
bandwidth of the system is finite and that constrains the symbol rate,
but it does not place a theoretical limit on the information-carrying
capacity (bit rate). Consider DTT transmissions, in which you can
increase the bit-rate by a factor of 1.5 by changing from 16-QAM (4
bits-per-symbol) to 64-QAM (6 bits-per-symbol) modulation, without
increasing the bandwidth at all.

The price you pay in using these higher-order modulation schemes is
sensitivity to noise. 64-QAM modulation will fail at a lower level of
added noise than 16-QAM will. But suggesting that the finite
bandwidth of the channel places an upper limit of the information-
carrying capacity of that channel (without saying anything about
noise) is plain wrong.

Richard.
http://www.rtrussell.co.uk/