How can digital be more spectrum efficient than analog ?

Less than half that actually. You must be counting responses too. In fact his total posts, to *all* groups, ever, show up as only 2060. (Not that those figures have any meaning, good or bad, except in your mind.)

Learn to use Google. And learn to search in groups where the topic is pertinent.

(If these two posts are any indication of the level of discussion/debate that goes on in comp.dsp, it is no wonder that I don't read that newsgroup.)

--
Floyd L. Davidson            
Ukpeagvik (Barrow, Alaska)                         floyd@apaflo.com

(snip)

By including part of the posting when you respond, like this :=)

Regarding the energy / power analogy, another failing is that there is no corresponding conservation laws. When information flows from source to 'losd' it remains undiminished at the source.

Also, dividing information by c^2 does not give you anything meaningful :=)

John

For those who know what they're doing and really understand how Shannon's analysis applies to communications, it's obvious what he meant by "forward-error-correction bits", at least it is to me.

And since it's pretty apparent that it's over your head I'll just reiterate that you have a lot of study left to do before you should claim to know better than the other folks posting in this thread.

No, real live study of a real live topic to the point that you grok the material.

So what did you mean by, "I did want to know what he meant by "forward-error-correction bits" though..."?

So I exaggerated. And my Mom told me a million times to never do that.

If I thought it would have done any good I would have, but it seems apparent to me that you're not interested in having a discussion where you might learn something. It should be readily apparent why someone might not care to waste their time on such an endeavor.

Eric Jacobsen Minister of Algorithms, Intel Corp. My opinions may not be Intel's opinions.

Don't you get masses of information? :-)

Steve

There do exist physics interpretations that, at Planck dimension, nature itself might be digital, not analog. There are also theories which put the total information content of the knowable universe at a finite (but large) number of bits, thus not continuous at all (at least in the mathematical sense).

IMHO. YMMV.

--
rhn A.T nicholson d.0.t C-o-M

Here's one, what I think is an intuitive picture, of why one can get more bps out of a analog channel of a given bandwidth.

Assume you have an analog channel of only a few Hz in bandwidth, but it was noiseless and you had a billion-bit resolution DAC/ADC pair. Then, just by modulating a

1 Hz carrier, you might be able to transfer > It's common knowledge that digital technology gives more telephone

IMHO. YMMV. random-top-bottom-middle-posting is the best :-P

--
rhn A.T nicholson d.0.t C-o-M

Sure. In the absence of noise -- that is, given complete certainty, all things are possible. A short metal bar can be a wonderful storage device. Whip out your trusty micrometer, measure its length expressed in binary. divide into octets, et voila! you have the entire text in the Library of Congress. On the other hand, maybe it's not so easy.

Jerry

--
        "The rights of the best of men are secured only as the
        rights of the vilest and most abhorrent are protected."
            - Chief Justice Charles Evans Hughes, 1927
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

I think you're getting close to the issue, Ron. Essentially, what you're bringing out is that Shannon's capacity theorem,

C = W*log_2(SNR+1)

depends on bandwidth (W) *and* SNR, while analog transmission usually only considers bandwidth.

I would build on your post by adding the fact that an equivalent analog system could transmit only a few Hz wide modulation signal through this same channel. In essence, it can't take advantage of the SNR.

This is the second time I've seen source coding and channel coding confused in this thread.

While source coding does also contribute to the "digital advantage," I believe the spirit of the original question was a question of channel coding and modulation. I think we would get to the "crux" of the issue if we focused on this aspect.

By the way, it seems to me that it's been amazingly hard to quantify this. We call ourselves professionals and can't answer a simple question like this???... :)

--
%  Randy Yates                  % "I met someone who looks alot like you,
%% Fuquay-Varina, NC            %             she does the things you do, 
%%% 919-577-9882                %                     but she is an IBM."
%%%%            %        'Yours Truly, 2095', *Time*, ELO   
http://home.earthlink.net/~yatescr

I wasn't attempting to confuse them as much as to try and say that one can do the latter without the former.

IMHO. YMMV.

--
rhn A.T nicholson d.0.t C-o-M

That formula *is* for analog transmission (as well as for digital transmission), which dependant on *both* SNR and bandwidth.

He was talking about exactly that though, an analog channel that can take advantage of the high SNR.

The same is essentially true if you use a channel has only a few dB SNR, but has infinite bandwidth.

In either case a digital or an analog system can use that channel for high throughput if the right analog modulation or digital encoding scheme is used.

I'm not clear on what your terminology means. How do you differentiate "source coding" and "channel coding"?

--
Floyd L. Davidson            
Ukpeagvik (Barrow, Alaska)                         floyd@apaflo.com

Here's is what he said,

Indeed, "a digital system can function error-free at an SNR so low that analog systems cannot function at all..." but that is entirely irrelevant. To achieve this magical feat, it must do so at a reduced *information* rate (forward-error-correction bits, retransmissions, etc. don't count as information).

1) It is *not* irrelevant. It is indeed a major part of the the reason the entire telecommunication industry switched from analog to digital. 2) It is not magical, and does not *require* "a reduced *information* rate ..."

What does require a (relatively) reduced information rate is virtually error free operation when the SNR is *not* high enough to reduce errors (which does happen occasionally with satellite and fiber optic systems, for example).

It may be obvious to you what he meant, but if you don't recognize it was incorrect...

Why is it *you* didn't recognize the incorrect statements?

(Do you have any actual experience with operation of equipment that uses FEC in a noisy environment? Have you ever seen such equipment run with it turned off, for example? A very good example would be watching a satellite system function as it goes through sun outages, or near sun outages.)

Why is it *you* didn't recognize the incorrect statements?

As noted above, the correlation he suggested is not valid...

And we note that you have yet to actually indicate that you have a valid answer to *anything*, much less that you actually do know of errors in any post.

So you still don't have an answer for anything....

--
Floyd L. Davidson            
Ukpeagvik (Barrow, Alaska)                         floyd@apaflo.com

(snip)

Yes, it is a poor way to deliver power, but if the source impedance is fixed and you want to maximize load power by varying the load impedance with everything else fixed it is right.

One place this currently comes up is in PoE, power over ethernet. With a Cat 5 cable and 48V maximum source (I believe due to electrical code) if you want the most power out you match the source (cable) resistance.

-- glen

Quite possibly, Steve. Or, as Rupert Murdoch discovered, divide the information content of the evening news by c^2 and you get information for the masses :=)

John

(snip)

Yes.

Many years ago I remember going on a tour of the Shasta Dam power station, and being surprised to find that the run at 1200RPM, though after a while it made sense.

I was still surprised to find a few years ago at Grand Coulee Dam that theirs run at 72RPM. I still wonder if you can see any effects of that in the frequency spectrum. I presume they have to minimize the 1.2Hz and 3.6Hz components, as we would probably notice them.

-- glen

(snip)

Not exactly but very similar.

My favorite example of impedance matching is the ping pong paddle and baseball bat.

Which one hits a baseball farther?

Which one a ping pong ball?

-- glen

As usual with quantum mechanics, it depends on how you do the detection. If the uncertainty in the number of photons is large enough, you can't really say it is an integer.

-- glen

(snip)

Maybe, but I have heard arguments like that use to say that for either DSL or cable the device at the end is not a modem because the signal is digital and not modulated. It IS an analog signal and is modulated/demodulated. We are averaging over very large numbers of very small photons.

Quantum cryptography runs in the countable photon regime, the rest of us are in the analog world.

-- glen

Because many analog systems are inefficient users of the available bandwidth. There is a huge amount of redundancy in an analog television signal to make it easier to build television sets. Removing that redundancy takes a huge amount of computing power, which we now have easily and cheaply.

-- glen

If it is an encoded digital signal, then technically it isn't really a modem. However... we've been refering to such devices as modems since they first existed. For example, T1 digital interface devices were (and are) often labeled as a "modem", even though it really was (is) just an encoder/decoder.

Of course if it actually is a mod/demod (cable modems, for example), then it truly is a modem.

I'm not sure which signal you are referencing there. Sensors in digital cameras, for example, fit your description and yes they very definitely are analog devices. On the other hand, the DSL modems that I'm familiar with are not analog devices, they use digital signals.

I'm not familiar with quantum cryptography, and other than "countable photon regime", I can't verify that it is digital. But yes, I obviously agree that typical photo measuring devices (e.g., those in a typical DSLR camera, are analog devices).

--
Floyd L. Davidson            
Ukpeagvik (Barrow, Alaska)                         floyd@apaflo.com