Questions about equivalents of audio/video and digital/analog.

[Floyd L. Davidson]

Funny, I don't see the word "digital" in there. Perhaps

You did see "discrete, unique value" right. Bingo.

you could point it out? No one is arguing that
"quantized" does not mean the above - but you seem
to be arguing that "quantized" is precisely equivalent
to "digital," while none of the definitions you provide
say that.

For people who cannot reason...

 

[Floyd L. Davidson]

(snip)

The pulse for which you provided the link, is not DSX-1, because it will not
fit within the DSX-1 envelope.

Really? Has the standard changed?

Or are you just missing something....

 

[Floyd L. Davidson]

Definitions are arbitrary only to those who don't
truly understand them.

You don't understand the word arbitrary either. In
post I read a bit ago you misunderstood the term "virtual"
too.

You have a lot trouble with words, don't you.

 

[Bob Myers]

That is not true. See the definition of analog data and
digital data.

The definitions according to yourself and your
selected "authorities"? Yet again? How about some
reasoning, instead, for a change?

Wrong.

See Shannon, "A Mathematical Theory of Communication,"
section V (27). Have someone help you with the big words,
as needed. When you're finished, find us an example of
a noiseless channel, and demonstrate to us what you're
saying.

Shannon discussed the theory of digital channels with
and without noise.

Shannon never used the word "digital," though, and used
the term "bits" simply because that is the commonly-used
unit of information (in ANY system), per information and
communications theory as it was established at the time
(and is still in use today). Given that you've admitted to
not being an authority in the field, I wouldn't expect that
you understood that before, so I'm more than happy to
give you this chance to learn. You're welcome.

He defined it as "a discrete channel
will mean a system whereby a sequence of choices from a
finite set ... can be transmitted ..."

Right - he defined a DISCRETE channel. The only
tie between "discrete" and "digital" appears to exist solely
in your own mind, since your cherished definitions of
"quantized" also unfortunately neglect to use the word
"digital" at any time. Shannon did NOT refer to either
"digital" or "analog" channels at all.

He then discussed the theory of analog channels, which
began with, "We now consider the case where the
signals or the messages or both are continuously
variable, in contrast with the discrete nature assumed
heretofore."

Wrong again. He discussed the theory of CONTINUOUS
channels, which is how they are consistently referenced in
his paper. In fact, the term "analog/analogue" does not appear
even once in Shannon's paper. "Analogous" appears a grand
total of six times, each time with the clearly-accepted meaning
of "similar to," as opposed to refering to any specific class of
signals. Similarly, "digital" does not appear at all. Shannon
correctly did not make any distinction between "analog" and
"digital" encoding in terms of information capacity or content,
as his theorems apply to any and all systems.

Clearly his definitions were consistent with those that
have been cited. Of course that again is just another
horrible appeal to authority... ;-)

Or it would have been, had Shannon actually provided
any such definitions. Too bad he didn't.

Bob M.

 

[Bob Myers]

You cannot have have not.

I'll wait to respond until that can be re-stated in something
resembling standard English.

Bob M.

 

[Bob Myers]

You did see "discrete, unique value" right. Bingo.

So "digital" = "discrete," as long as we accept your
claim, not backed by ANY definition or reasoning,
that it does?

For people who cannot reason...

Oh, NOW it's not about the definitions, it's about
"reason," eh? Fine - reason away. It will be, at the
very least, a refreshing change.

Did you learn this flip-flopping technique ("It's all
about the definitions!" "No, it's not actually IN the
definitions, it's all about reason!") from any
politicians you'd care to name?

Bob M.

 

[Bob Myers]

There's a difference between quantization and quantification. A sample and
hold, for example, quantizes an analog signal. It is considered quantized
because a signal might now be represented as a series of sampled levels
rather than a continuously varying level. But the held samples are not
digitized until they have been quantified, i.e. a number has been assigned
to each sample that is ratiometric with the analog quantity being
digitized. In the digital domain, the signal is comprised of a series of
numbers each of which represents a discrete (quantized) analog value.

Oh, gee, Jim, you obviously just don't get it, either, or
so I am expecting Floyd to claim. Gee, and you
SOUNDED like an "authority," too....

The above was very nicely stated, by the way.

Bob M.

 

[Arny Krueger]

OK, my apologies; I was under the impression that you
considered the line ITSELF to be "analog."

Well, I would agree that they carry *electrical* signals; they
are not necessarily either "analog" or "digital" unless
someone (whoever THAT might be) insists on using "analog"
to mean "continuous," when there's already a perfectly good
word at hand (for instance, uh...."continuous!") that can be
used for that.

The reason I say that they are analog at the most fundamental level is the
fact that every line receiver I've ever seen makes some kind of
analog-domain judgement about the state of the line. Like, 0 is less than
0.5 volts and 1 is greater than 2.7 volts.

 

[Floyd L. Davidson]

Arny Krueger wrote:

(snip)


I don't believe V.90 works that way. That is, you can't
connect two V.90 modems together that way.

You cannot connect two subscriber modems end to end and
use v.90 protocols.

But you can use to provider end modems and get v.90
protocols running in both direction, and in fact if the
line is good enough it can be made to run a 64 Kbps.

The answer end of a V.90 connection must be an ISDN line, usually a

It must be digital. ISDN is one possibility, but not the only
one.

primary rate line with 24 channels. The result is that there is only
one A/D and D/A conversion between the two end points.

Exactly. The "A" part of that is a digital PAM signal.
If it goes through and extra CODEC, the data is
scrambled.

You may say that an ISDN line is an analog line since you can measure
the voltage on the wire as a function of time. I won't try to argue
either way on that one.

That would be an absurd claim.

ISDN is Integrated Services Digital Network. It is a
digital signal in all respects, using what is called
2B1Q encoding on a customer Basic Rate Interface loop.
That is 2 Binary 1 Quaternary meaning there are two
binary bits encoded in each symbol using 4 possible
voltage levels.

 

[Floyd L. Davidson]

Apparenlty Claude Shannon didn't agree with you on that.

And apparently you don't understand what was meant in
the above.

It is theoretically impossible for any real-world communications
channel to be noise-free or possessed of infinite bandwidth.
Do you disagree with this statement? If so, please show the
error.

This does not prevent a noise-free channel from being IMAGINED,
and used as the basis for a mathematical analysis, which is
what Shannon did. But Shannon most definitely did NOT
expect any such thing to be realized, and fully understood
why it could not be.

Have you even read Shannon's paper? In section V (27),
Shannon makes virtually the same statement I gave earlier
re the notion of "infinite" levels:

"This means that to transmit the output of a continuous
source with *exact recovery* [emphasis Shannon's]
at the receiving point requires, in general, a channel of
infinite capacity (in bits per second). Since ordinarily
channels have a certain amount of noise, and therefore
a finite capacity, exact transmission is impossible."

Gee, here's another puzzler for you - throughout his
paper, Shannon discusses channel capacity in terms of
"bits per second." Does this mean that his work is
applicable only to digital systems? If not, why not?

So you have show that the in practice part is true, and
I have show that the in theory part was wrong. A noiseless
channel can exist in theory (not in practice). It is
a very usefule theory to study too.

 

[Don Bowey]

At some level every signal is analog. The digital nature of a signal comes
from how it is coded and decoded.

Well! To borrow a query method from you; ????????????????????

 

[glen herrmannsfeldt]

Arny Krueger wrote:

(snip)

The absence of digital coding, for one thing. Back in the old days, all
lines were analog. I think that maybe the telegraph put an end to that.
Morse code is a sort of PWM, right? ;-)

The telegraph and digital communication came first.

http://en.wikipedia.org/wiki/Telegraph

Later, voice was sent over telegraph lines, the conversion to
analog signaling.

http://en.wikipedia.org/wiki/Telephone

Note again that Nyquist was figuring out how fast he could send
telegraph pulses through an analog (band limited) line.
(Underwater cables are especially limited by capacitance, when
compared to air insulated open wire transmission.)

-- glen

 

[timepixdc]

It is theoretically impossible for any real-world communications
channel to be noise-free or possessed of infinite bandwidth.
Do you disagree with this statement?

It would appear that you disagree with your own statement since further
down you write,

This does not prevent a noise-free channel from being IMAGINED

Theory? Imagination? Can't have the first without the second.

X-no archive

 

[Bob Myers]

It would appear that you disagree with your own statement since further
down you write,

What part of "real-world" (as opposed to "imaginary")
in the above did you fail to comprehend?

Bob M.

 

[Don Bowey]

So you don't know of any telco that uses them today.

I assume you were also using mechanical switching there
too... ;-)

It is sort of difficult for me to imagine that sort of
environment, as Alaska was fully digital when the rest
of the country had only gone 33% digital. By the mid-1980
the only mechanical switches left in Alaska were owned by
the military, and they were gone by 1990.

Still, I don't think anyone *ever* used E type repeaters
in Alaska, but I could be wrong on that.


The general design paradigm used now is to put
"remotes" at multiple strategic sites and control
them all from one digital switch. Of course all of
these are trunked together, and the whole idea is to
prevent long loops while also requiring administration
of only a single digital switch.

That was a basic design decision made for telco's by the
vendors, back in the late 80's or early 90's. It was
enforced with system pricing! Nortel (NTI at the time),
for example, simply made the software for a digital
switch (actually, the long term use and maintenance of
the software) far more expensive than installing
remotes. It became uneconomical to have two switches in
any jurisdiction where it was possible to deactivate one
and replace it and move forward with remotes.

By the mid-1990's all of NTI's customer base had moved
in that direction.


My concern was only the State of Alaska... which is
of course the size of 20% of the entire Lower-48.

Gosh, after all that I guess I should be impressed.......... But........ I'm
not. I wonder why.

 

[Don Bowey]

I've given you the URL for a glossary of terms, why
don't you use it? Here are some others:

http://www.atis.org/tg2k/
http://www.itu.int/sancho/index.asp
http://www.carrieraccessbilling.com/telecommunications-glossary-a.asp
http://www.faxswitch.com/Definitions/


Quasi-Analog Signal -

A digital signal that has been converted to a form
suitable for transmission over a specified analog
channel.

Gobbledegook.

 

[glen herrmannsfeldt]

Bob Myers wrote:

(snip on analog vs. digital)

Well, I would agree that they carry *electrical* signals; they
are not necessarily either "analog" or "digital" unless
someone (whoever THAT might be) insists on using "analog"
to mean "continuous," when there's already a perfectly good
word at hand (for instance, uh...."continuous!") that can be
used for that.

This reminds me of the discussions I used to see along the
lines of (technology X) is digital and not analog,
the signal is not modulated, such that
(device used to connect to X) is not a modem.

In most cases, the signal is modulated, and so the device
can reasonably be called a modem. The two that this has
come up most often are for ISDN and DSL.

Note that even baseband ethernet uses a modulation system:
synchronous phase modulation (also known as Manchester
coding). This modulation is necessary to be able to
extract a clock to recover the signal.

I might agree that NRZ, and maybe NRZI, are unmodulated
such that the connecting device isn't a modem and the signal
on the line is 'digital'. Because of the need for a separate
clock, NRZ is rarely used for longer distances.

-- glen

 

[Don Bowey]

Arny Krueger wrote:

(snip)


I don't believe V.90 works that way. That is, you can't connect two
V.90 modems together that way.

The answer end of a V.90 connection must be an ISDN line, usually a
primary rate line with 24 channels.

That is nonsense. ISDN has nothing to do with it. Any modern digital
message trunk provides an identical channelizing process.

The result is that there is only
one A/D and D/A conversion between the two end points.

Same as the Message Network.

You may say that an ISDN line is an analog line since you can measure
the voltage on the wire as a function of time.

What are you calling an ISDN line? The ds1 rate cable pairs? DS3
transport, Sonet, the analog channel that pops out of the terminal
multiplexer?

 

[Don Bowey]

Really? Has the standard changed?

Not that I know of. ANSI T1.403 does good. Want to argue with it?

Or are you just missing something....

I don't believe so. On the other hand, I inquired to you about the DSX-1
envelope, and you replied with a link to a waveform you drew, which has only
a loose connection to DSX-1. It appears you might be missing something.

 

[Don Bowey]

You cannot connect two subscriber modems end to end and
use v.90 protocols.

But you can use to provider end modems and get v.90
protocols running in both direction, and in fact if the
line is good enough it can be made to run a 64 Kbps.


It must be digital. ISDN is one possibility, but not the only
one.


Exactly. The "A" part of that is a digital PAM signal.
If it goes through and extra CODEC, the data is
scrambled.


That would be an absurd claim.

ISDN is Integrated Services Digital Network. It is a
digital signal in all respects, using what is called
2B1Q encoding on a customer Basic Rate Interface loop.
That is 2 Binary 1 Quaternary meaning there are two
binary bits encoded in each symbol using 4 possible
voltage levels.

Your reply is incomplete.

* "ISDN is Integrated Services Digital Network." Right

* "It is a digital signal in all respects...." Right

* "...using what is called 2B1Q encoding on a customer Basic Rate Interface
loop. That is 2 Binary 1 Quaternary meaning there are two binary bits
encoded in each symbol using 4 possible voltage levels." Incomplete,
therefore wrong.

There is also a Primary Rate ISDN, which is a DS1 rate and which uses the
AMI line code, modified by B8ZS. One end of a Primary Rate channel may be in
a Central Office or both ends may be at a customer premises.