«

On 5/31/11 6:35 AM, Whisky-dave wrote:

> Which is an important point in the real world.
> In the UK we are getting rid of our analogue TV and replacing it with
> digital
> because the amount of 'air space' analogue takes up is far higher than
> digital.
> So in that 'transmission space' we get 5 analogue channels of crap
> or we can get ~100 channels of digital crap... a bargain if ever I saw
> one

Is there a change in the resolution with the switch to digital TV
transmission? Still using PAL? Is it HD capable (1920 x 1080 pixels)?

I was always envious of Brit TV until we got High Definition TV in the
States.

Eric Stevens <> wrote:
> On Thu, 26 May 2011 22:26:09 -0700, nospam <>

>>> A top quality vinyl record played with a top quality moving coil
>>> cartridge connected to a top quality (not not necessarily fancy)
>>> amplifier and top quality speakers not aimed at heavy metal will
>>> extract music beyond the ability of the
>>> _best_commercially_available_digital_.

>>bullshit. that's simply not possible. digital will have lower noise,
>>wider dynamic range, no wow&flutter, no pops&clicks and much lower
>>distortion.

> You are not necessarily correct about the distortion, especially when
> you realise the last part of the chain before the speakers is a
> non-digital amplifier.

The same non-digital amplifier that's the last part of the
chain before the speakers when you use vinyl?

-Wolfgang

Mxsmanic <> wrote:
> Whisky-dave writes:

>> The number of eggs a chicken lays is digital whether it laid them or
>> not before ...

> A number is not physical. It's just a mental concept.

A number of eggs is very physical. If you disbelieve, let me
throw them at you.

>> ... was there to hear it is yet another question and a digital one.

> Questions are not physical entities, either.

Question are interfaces to other people or nature etc. All
interfaces are analog.

>> No, most monetary transactions are digital, I buy six eggs for a set
>> price all is digital.

> Transactions are not physical entities.

A coin is a symbol, a token, signifying a certain value.

-Wolfgang

"John McWilliams" <> wrote in message
news:is3au4$glm$...
> On 5/31/11 6:35 AM, Whisky-dave wrote:
>
>> Which is an important point in the real world.
>> In the UK we are getting rid of our analogue TV and replacing it with
>> digital
>> because the amount of 'air space' analogue takes up is far higher than
>> digital.
>> So in that 'transmission space' we get 5 analogue channels of crap
>> or we can get ~100 channels of digital crap... a bargain if ever I saw
>> one
>
> Is there a change in the resolution with the switch to digital TV
> transmission? Still using PAL? Is it HD capable (1920 x 1080 pixels)?
>
> I was always envious of Brit TV until we got High Definition TV in the
> States.
>
>

Yes, HD TV here in the US has been great for me. In the past, when a storm
was headed to the area and I wanted to know if a tornado was going to hit, I
could get five snowy, but watchable stations. Now, when a tornado is bearing
down on the area, I got nothing. HD signals are either receivable or not, no
middle ground.

OTOH, I can watch "court" (Judge Judy, People's Court, etc) shows for six
straight hours every weekday afternoon. And on Saturdays, there are
infomercials all over the dial.

On Tue, 31 May 2011 08:19:31 -0800, Floyd L. Davidson wrote:
[Quoting some unnamed standards organization, perhaps even the ISO.]
> Once again, here are the *standard* definition. Every standards
> organization in the world uses exactly the same definitions. This is
> not a matter of opinion, perspective, or whatever. It is the
> *definition* of the distinction between analog and digital data amd
> signals.
>
> The key words to note are continuous vs discrete, and an infinite set as
> opposed to a finite set, of symbols. Those are the essential
> differences, and *anything that does not comply cannot be correct*.
>
>
> ANALOG DATA:
>
> Data represented by a physical quantity that is considered to be
> continuously variable and has a magnitude directly proportional to
> the data or to a suitable function of the data.

I consider voltage to be a physical quantity that is continuously
variable, and in the case of D/A output it happens to have a magnitude
directly proportional to the data (or a function thereof.) It just
happens that in this particular instance the data itself was discrete,
quantized.

> DIGITAL DATA:
>
> 1. Data represented by discrete values or conditions,
> as opposed to analog data.
>
> 2. Discrete representations of quantized values of
> variables, e.g. , the representation of numbers by digits,
> perhaps with special characters and the "space" character.
>
>
> ANALOG SIGNAL:
>
> 1. A signal that has a continuous nature rather
> than a pulsed or discrete nature.
>
> Note: Electrical or physical analogies, such as
> continuously varying voltages, frequencies, or phases, may
> be used as analog signals.

A unit-hold amplitude modulated voltage isn't "pulsed" in nature, and its
discreteness is only a function of the data that it is an analog of.

> 2. A nominally continuous electrical signal that
> varies in some direct correlation with another signal impressed
> on a transducer.
>
> Note: For example, an analog signal may vary in
> frequency, phase, or amplitude in response to changes in
> physical phenomena, such as sound, light, heat, position, or
> pressure.
>
>
> DIGITAL SIGNAL:
>
> A signal in which discrete steps are used to represent information.
>
> Note 1: In a digital signal, the discrete steps may be
> further characterized by signal elements, such as
> significant conditions, significant instants, and
> transitions.
>
> Note 2: Digital signals contain m-ary significant conditions.

That's got to make a modern, dynamically channel-equalized communication
system with a multi-bit modulation scheme "analog", then: nothing
discrete or step-like going on there...

I don't think that the "definitions" that you're relying on are useful
(or even unambiguous.)

Cheers,

--
Andrew

On Tue, 31 May 2011 16:45:31 -0800, Floyd L. Davidson wrote:

> Eh? I'm not sure what you mean by a "modern, dynamically
> channel-equalized communications system", or how those particular
> adjectives are significant to your conclusion.

The signals produced and consumed by modern communication systems are
about as continuously changing, continuous in level and variable in
nature as it is possible to get, because (a) the channel is continuously
changing in an analog, continuous fashion and (b) that's how to maximise
the bandwidth of the channel.

Old-school digital communication systems like RS-232 or 10base2 ethernet
had a direct two- or three- level system with a straight-forward
relationship between the bits and the levels, but those aren't the
popular communication standards any more. OFDM wireless and other
contemporary systems aren't so easy to categorize.

> By definition (again, you do not have the option of redefining these
> terms), if it is a "multi-bit modulation scheme" it *necessarily* is
> made up of discrete values. Each bit *is* discrete, and there is no
> continuity between two bit values, which by definition means it *cannot*
> be analog.

The "bits" are discrete, but they are the result of sophisticated
filtering and statistical decision systems built into the receivers. The
voltages or EM field strengths in the channel appear continuous (indeed
indistinguishable from full-scale white noise) to anyone who was to
observe them on an oscilloscope or other similar instrument.

[I apologise to any readers actually interested in viewfinders. I hope
and expect that this will be my last post on this off-topic topic.]

Cheers,

--
Andrew

On May 31, 2:45*am, fl...@apaflo.com (Floyd L. Davidson) wrote:
> Noons <wizofo...@yahoo.com.au> wrote:
> >Not really. *I think the problem is with the wording. *For digital and assuming
> >binary, you must have two arbitrary levels of analog signal to encode the 1 and
>
> So like the person who made the orginal statement this
> all references, you too are going to confuse "digital"
> with "binary".

Is that what in whatever deranged language you speak, "for digital and
assuming binary" means? Because, rest assured: it is only in that
language.

> *All binary encoding is necessarily
> digital, but *most of the the digital world is not
> encoded as binary*.

That is why I said "assuming binary". You really make a special effort
to show your inability to read common English, don't you?


> Most digital channels (all of which are "transmitted
> across an analog channel")

No. Most use MANY analog channels. Not a single one.

> use m-ary encoding where m is
> greater than 2. *Base 16 is no more common than say 256
> or 64, but binary certainly is not more common either.

Amazing. And your point is?.....

On May 31, 8:40*am, Eric Stevens <eric.stev...@sum.co.nz> wrote:



> I accept your point. What I had in mind was the contamination of a
> digital signal (of the kind you have described) where either 1 gets
> lost or a 0 is filled with a pulse of external noise. Either way the
> signal will contain an error.

OK, got it. That's why most digital transmissions use some form of
redundancy.
That can take the form of error correction and/or parity bits when
we're talking binary or its multiples. The problem being of course
that there is no such thing as "digital" transmission: ALL signal
transmissions use an analog base. It's how we encode/decode that
analog signal from/to digital that makes the difference.

Shannon's stuff gave us a mathematical base to determine exactly how
much redundancy do we need, depending on a number of factors which
include but are not limited to the analog bandwidth.

But of course it said nothing about the analog media to be used: that
is entirely the domain of another engineering discipline. And it said
even less on how do we get from pure analog to digital in the first
place via sampling, which is the implication Floyd keeps making and is
completely, utterly wrong.

The bottom line for transmission is the term "redundancy". If that is
achieved through additional digital encoding or through extended
bandwidth of the analog media, that is entirely determined by how much
noise is acceptable and can be/must be avoided.

If the analog channel allows for higher usable bandwidth - for
example, optical fibre transmissions - then the noise cancelling
through digital encoding may be reduced. Note: I said "may", I didn't
say "has to". If on the other hand the analog bandwidth is very small
- for example, the vertical magnetic recording on modern disks - then
the compensation in error recovery on the digital side has to be
increased: the likelyhood of errors is much increased.

Which was my interpretation of what the statement meant and why I
agreed with it.


> There are very few forms of data transmission where a simple binary
> stream is sufficient. In virtually all cases the binary stream will
> encode something more complex.. From the early days of digital
> communication it is common to code data to base 8 (Octal) or base 16
> (Hexadecimal). Either way, a single data item consists of a train of
> pulses and non-pulses from which the value of the data item can be
> decoded. A missing pulse or a spurious pulse will change the value of
> the decoded train.

Agreed.


> This is where we are at cross purposes. The statement " Anything below
> the threshold is treated as noise, even if it's signal, and anything
> about(sic) the threshold is treated as signal, even if it's noise" is
> applied to the simple binary signal on which the data is encoded. I
> might have written that as " Anything below the threshold is treated
> as no signal, even if it's signal, and anything above the threshold is
> treated as signal, even if it's noise".

But that is the problem: analog encoding of digital uses two arbitrary
levels, a lo and a high one within an arbitrary signal range.
Anything outside those two boundaries is noise, be it under or over.
An overload is not a "1". It is noise. Same for an underload.

The device used to detect a digital signal transition (do not confuse
with an analog-digital conversion: this is digital domain only) is in
almost all cases an amplifier which detects a transition between two
levels of analog stimulus.
In operational amplifier analog terms, such a device is called a
diferentiator: its output swings between two extremes *only* (by
convention the 0 and 1) as the input moves up and down the analog
signal range.

Any transition between any two analog levels outside of the arbitrary
signal range that triggers that transition is not detected as such and
hence there is signal loss. That means an underload causes as much
signal loss as an overload: what is missing is the transition through
the range defined as the signal range.



> Seeing we comparing analogue with digital I assumed the binary signal
> under discussion encoded a higher level of data for which it is
> inappropriate to say that an arbitrary value determines the difference
> between noise and signal.

I think the problem here is that a digital message is completely
defined by the chosen encoding. There is no such thing as a "higher
level of data". Eg: one byte is one byte, it can only encode 256
different values. It doesn't matter how high or low it is encoded in
an analog transmission, it remains a byte with only 256 combinations
of 8 bit values possible.

Of course, by *convention* you can attribute a given decimal value to
the binary encoding of 00000000, which happens to not be 0. And
another to 11111111 that is not 256. But that is a totallly different
thing, although quite common: that's how dynamic range compression
happens, for example.

Interesting subject. Thanks for the normal discussion and not snipping
and changing randomly, like Floyd does.

On May 30, 11:03*pm, Walter Banks <wal...@bytecraft.com> wrote:


> Essentially what you are saying is correct about information theory.

Thank you. It was part of my electrical engineering course 40 years
ago. And the one that I got the best marks on, hence why I ended up in
IT.

> What is different with camera's in the last few years is the sensors in
> the better digital camera's have reached the point where their dynamic range
> is better than the analog dynamic range of film with a lower signal to noise.

Not really. What happened is that manufacturers finally got enough
processing power jammed into the cameras to be able to assign and
calculate on the fly different levels of encoding, resulting in the
equivalent of a dynamic range compression effect.

You can do exactly the same with film:in fact any colour negative film
does it as a matter of routine and the entire zone system for b&w is
based on it!

It's how you manipulate that compression that creates the challenge
with film. With digital sensors, it diferentiates for example the
results in Sony's sensors between Sony cameras and Nikon cameras: both
use the same sensors, each processes the data differently.


> Digital post processing technology has developed significantly as well
> capable of producing outstanding results.

Absolutely, 100% in agreement! Hence why my film scans nowadays are
incredibly better than ever before, even though I am still using the
same scanning technology.

> Film in some forms is likely to remain for years increasingly becoming an
> art form.

Thank the Gods for that! It's exactly what it always was and will
remain, hopefully.

On May 31, 8:55*am, Eric Stevens <eric.stev...@sum.co.nz> wrote:


> Your snoot seems to be upset.

Sorry, I was upset by something else. My apologies.


> "In physics,

that essentially is the problem: "in physics". Thats is not
necessarily the same as "in nature".

A photon is a physics abstraction: it has not been proven anywhere
that it exists, no more than a light wave does.


> I wasn't going to try and unravel all that for the sake of this
> argument. I used the word 'photon' in the simplest way possible in the
> hope that it would be understood by all. I seem to have failed.

Not at all. You just forgot to recall or note that you were talking
in the universe of physics. Which as we all know has different and
multiple dimensions - 11 by last count!


> Being created by the accumulation of small numbers of individual
> electrons, the electrical charges are indeed quantized. It is not
> possible to have an electrical charge less than that carried by
> electron.


Yes, but once again we are mixing things. An singular electrical
charge is *not* the same as a singular light charge. NO one has
proven yet that there is such a thing as a photon, although for
convenience that abstraction can be made in the field of physics.