[William Sommerwerck]

For goodness sake - how are you not getting this? If the error in
length is less than your quantization step, then it simply isn't an
error.

Correct.

If it is a much larger error (which is far more likely) then it
doesn't matter whether there is quantization or not - the result is
the same.

If that occurs, it's because the system is poorly designed. A "good" digital
system does not have large numbers of errors that exceed the bounds that
would prevent correct reading.


The reason digital works is because you store the numbers, not some
variable quantity that represents their magnitude.

You need to think this through. Most people -- including those with college
degrees -- just don't understand that any quantization of sample amplitude
is digitization of the signal.