In article , Bob Latham wrote:
You'll effect these things by reducing the data rate. But no cable can
do do this.

I'm almost in full agreement with this but due to lack of knowledge there
are still a couple of "maybes" in my head.

A cable carrying a digital data stream presumably square waves, will
distort the shape of that wave. Hope we have no argument so far.

Not actually square waves because that would require infinite bandwidth, but
a defined waveshape, which if necessary can be recreated by resampling. Yes,
the waveshape can become distorted, but provided resampling is done before
distortion is bad enough to cause errors, the error rate due to distortion in
the transmission path itself can be kept to zero.

If I knew that the receiver was correctly managing to resolve the ones and
zeros and that their edge shape was having no effect on the DAC because
there was a 1 kilo byte or so FIFO buffer before the DAC and that no
additional "guessing" due to distorted data was going on then that would
be great.

That's the whole point of digital transmission. Details vary, but the general
principle with one-way transmission is that slightly more information is sent
than would be necessary to encode the signal if transmission were perfect.
The extra information is encoded in such a way that random errors below a
certain error rate can be completely corrected, i.e. there is suffcient
information to recreate the exact original bitstream.

Also, individual bytes of the information are redistributed in such a way
that a data dropout (a burst of interference on a radio transmission, or a
blemish on a disc or tape surface) will be very unlikely to obliterate
samples that were adjacent in the original bitstream. Therefore, even at an
eror rate higher than that at which perfect error correction can be done, it
is still possible to make a guess at the missing samples by interpolation
because the ones immediately on either side of them will be intact. We might
call this "error concealment", rather than "error correction". While it
wouldn't be possible with a computer file where every byte must be perfect,
it works very well where the bytes represent samples of partially predictable
material such as audio or video.

To summarise, if proper care is taken with transmission/recording of digital
audio and video signals, errors can be kept to zero even allowing for some
random dropouts due to interference or recording surface blemishes, and even
some dropouts in excess of what can be corected can be convincingly
concealed.

Couple that with knowing that there was no effect being caused by the
quality (or lack of) of ground connection between the two devices and I
would then be happy to say the above is 100%. The cable is having no
effect and anything heard is imagination.

There's the Q-word again. What physical property are you talking about when
you refer to the "quality" of a ground connection? If it's something real, it
can be measured and kept within defined limits. Engineering isn't magic.

I don't have that knowledge.

I hope I have managed to supply some that is helpful.

Rod.