Back to basics. Definition: Analogue audio

[Ken Tajalli]

Of course not.

Good, we are here to learn after all.

It is just somewhat obvious that a digital signal on an interconnect (wire/cable/fibre/pcb trace etc) is transmitted with analogue signals (values of voltage/current).

Is it? By definitions offered so far, and according to Wikipedia and other sources, an analogue transmission is a smooth varying signal vs time, as opposed to pulses and codes. Does a digital signal going through wire PCB or any other medium has undefined, unlimited voltage levels?
Everybody knows, it's obvious .... etc. won't do - can you define precisely what analogue transmission is? And then does digital transmission get covered in that?

There is no such thing as a square wave, they have rise and fall times. The rise and fall are not straight lines, the "top" and "bottom" of the "square" wave are not flat. The transitions from rising/falling edge to top/bottom are not sharp corners.

What does square wave got anything to do with it? Digital transmission can be modulated in radio waves.

In fact digital waveforms can (especially at higher speed) look surprisingly close to sine waves.

.... and yet can perfectly be decoded.
There you go, you answered yourself.

In other words, they are continuously varying voltages - different from one moment to the next.

Perhaps, but does the variation in those voltages carry any encoded information? Or are they by-products of the transmission line inadequacies? Can the digital receiver decode the original data, despite distortions and the noise?

These are the questions I have encountered.
Just give up, you will not win, it is obvious , ...... won't do.

 

[pkane]

Good, we are here to learn after all.

Is it? By definitions offered so far, and according to Wikipedia and other sources, an analogue transmission is a smooth varying signal vs time, as opposed to pulses and codes. Does a digital signal going through wire PCB or any other medium has undefined, unlimited voltage levels?
Everybody knows, it's obvious .... etc. won't do - can you define precisely what analogue transmission is? And then does digital transmission get covered in that?

What does square wave got anything to do with it? Digital transmission can be modulated in radio waves.

.... and yet can perfectly be decoded.
There you go, you answered yourself.

Perhaps, but does the variation in those voltages carry any encoded information? Or are they by-products of the transmission line inadequacies? Can the digital receiver decode the original data, despite distortions and the noise?

These are the questions I have encountered.
Just give up, you will not win, it is obvious , ...... won't do.

All transmission uses analog signals unless you get down to individual elementary particles. I don't know what the debate is about. How the analog signal is interpreted is either digital or analog.

 

[MRC01]

All transmission uses analog signals unless you get down to individual elementary particles. I don't know what the debate is about. How the analog signal is interpreted is either digital or analog.

Yep. At the scale where we perceive and manipulate the universe, it is analog/continuous. A perfectly discrete physical state implies instantaneous change (from before that state, and after that state), which implies infinite rate of change, which implies infinite energy/power, which doesn't exist. At least not at the super/macro-atomic scale at which we perceive and manipulate the universe.

So the only means we have to physically encode (store or transmit) information, manipulating the universe around us, must necessarily be analog/continuous.

But, we can interpret, translate that physical medium as a set of fully discrete states, in both directions (discrete information to analog/continuous physical encoding, and back).

This feels a bit like debating how many angels can fit on the head of a pin.

 

[Ken Tajalli]

All transmission uses analog signals unless you get down to individual elementary particles. I don't know what the debate is about. How the analog signal is interpreted is either digital or analog.

Have any reasons for that?
I did ask a few questions, I assumed you answer those in a scientific manner, not just " it is, because I said so"
Please reason it for me.
EG I say a continuous DC current in a circuit is neither digital signal nor analogue signal, as it carries no information but its existence, which is a very simple binary data.
For it to be a signal, specially in audio, it needs to carry complex uncertain information.
An analogue data transmission has no discrete levels, any changes in amplitude vs time is the information it carries.
In contrast, a digital data transmission does have discrete levels, and variations in amplitude vs time, save for expected discrete levels, do not constitute as information.
Whether they are carried electrically, optically or by RF carrier wave (such as DAB) is irrelevant.
A Morse-code, transmitted by a flashlight at night between two boats, is a digital transmission, small variances in brightness due to thin sea-mist are discarded so are small timing jitters.
Now if the sea-mist gets thick enough to block the light completely for any length of time or the sun should come up (noise gets as large as signal), then the receiver (depending on how long the light was blocked) either gets nonsense because it does not conform to Morse-code table and the language being used (large error) or no signal at all. Then he would signal repeat.

 

[antcollinet]

Good, we are here to learn after all.

Is it? By definitions offered so far, and according to Wikipedia and other sources, an analogue transmission is a smooth varying signal vs time, as opposed to pulses and codes. Does a digital signal going through wire PCB or any other medium has undefined, unlimited voltage levels?
Everybody knows, it's obvious .... etc. won't do - can you define precisely what analogue transmission is? And then does digital transmission get covered in that?

What does square wave got anything to do with it? Digital transmission can be modulated in radio waves.

.... and yet can perfectly be decoded.
There you go, you answered yourself.

Perhaps, but does the variation in those voltages carry any encoded information? Or are they by-products of the transmission line inadequacies? Can the digital receiver decode the original data, despite distortions and the noise?

These are the questions I have encountered.
Just give up, you will not win, it is obvious , ...... won't do.

It's not about winning or losing - it is about physics.

 

[Feelas]

Have any reasons for that?
I did ask a few questions, I assumed you answer those in a scientific manner, not just " it is, because I said so"
Please reason it for me.
EG I say a continuous DC current in a circuit is neither digital signal nor analogue signal, as it carries no information but its existence, which is a very simple binary data.
For it to be a signal, specially in audio, it needs to carry complex uncertain information.
An analogue data transmission has no discrete levels, any changes in amplitude vs time is the information it carries.
In contrast, a digital data transmission does have discrete levels, and variations in amplitude vs time, save for expected discrete levels, do not constitute as information.
Whether they are carried electrically, optically or by RF carrier wave (such as DAB) is irrelevant.
A Morse-code, transmitted by a flashlight at night between two boats, is a digital transmission, small variances in brightness due to thin sea-mist are discarded so are small timing jitters.
Now if the sea-mist gets thick enough to block the light completely for any length of time (noise gets as large as signal), then the receiver (depending on how long the light was blocked) either gets nonsense because it does not conform to Morse-code table and the language being used (large error) or no signal at all. Then he would signal repeat.

Well, would it work for a compact definition that an analog signal is the information, while a digital signal is information only when intepreted through a known schema? Whether a signal carries information or not is due to the receiver to decide. You can work with a continous DC current as an analogue audio signal. This is not a debate about analogue/digital sound, but merely about what "sound" is. Why would you choose not to treat continous DC current as not an analogue signal, if it can be played through speakers or measured on an oscilloscope? I'd say that DC carries a constant information, not a lack of information. Yet, perhaps, it feels intuitive to treat presence of a signal as an information on it's own. Either way, the guidelines to interpretation are always outside of the signal itself. Somebody brought out that continous DC current is time invariant, which does not imply non-continuous as you implied earlier.

Does a digital signal going through wire PCB or any other medium has undefined, unlimited voltage levels?

Of course it can have unlimited and undefined voltage levels, unless it was designed not to. Choice of levels is completely arbitrary when not working with any industry standard. Analogue -> digital is a translation after all.

Perhaps, but does the variation in those voltages carry any encoded information? Or are they by-products of the transmission line inadequacies? Can the digital receiver decode the original data, despite distortions and the noise?

Well, theoretically you could measure variation of those voltages, i.e. to understand what transmission line is being used. I feel like wrong questions are being asked here.

I would go as far as to say that signal is neither "analogue" nor "digital" and it's type is dependent entirely on whatever the receiver & sender agreed upon.

 

[Ken Tajalli]

Well, would it work for a compact definition that an analog signal is the information,

If it does not vary, there is no information being carried. If it varies in an uncertain way through indefinite levels, then Yes, it is analogue.
Also bear in mind, that it can be argued that a simple sine-wave is an analogue, which it is, but since we expect it to vary in a certain manner, it carries no information, but its existence, which again would be a simple binary. (such as AM radio carrier wave)

while a digital signal is information only when intepreted through a known schema?

Again, information is the uncertainty in the signal, meaning what we expect it to be there is no information, the handshake protocol, the timing pulses etc.

Whether a signal carries information or not is due to the receiver to decide. You can work with a continous DC current as an analog audio signal.

No audio or complex data can be carried over by a constant, continuous DC current in a circuit, but a simple binary, two state info.

It is getting late over here, I am gonna say good night.

 

[Feelas]

If it does not vary, there is no information being carried. If it varies in an uncertain way through indefinite levels, then Yes, it is analogue.
Also bear in mind, that it can be argued that a simple sine-wave is an analogue, which it is, but since we expect it to vary in a certain manner, it carries no information, but its existence, which again would be a simple binary.

Yet a simple sine-wave has cycles and can carry a simple audio signal, thus it seems intuitive to call it analogue audio. Frequency is always the information in sines. Why can't the information stay constant?

No audio or complex data can be carried over by a constant, continuous DC current in a circuit, but a simple binary, two state info.

You are cheating here!
You can carry audio through a simple binary, two state info if you put the audio in the receiver & make it play it or not, depending on DC presence. Either way the convention is outside of the system for digital and it isn't necessary in analogue, where you take the signal at it's face value.

 

[pkane]

Have any reasons for that?

Yes, Physics and information theory.

EG I say a continuous DC current in a circuit is neither digital signal nor analogue signal, as it carries no information but its existence, which is a very simple binary data.

Of course it carries information beyond its existence: the actual current, which is in no way binary. But, this is exactly my point: it's not the carrier that defines whether the signal is analog or digital. It is the interpreter. If I look at whether DC current exists or not, that's one bit of digital information. If I look at the current level, that's analog. Same carrier, two different interpretations, both can be affected by noise to a different degree.

Again, information is the uncertainty in the signal, meaning what we expect it to be there is no information, the handshake protocol, the timing pulses etc.

No. Information and uncertainty are most certainly not the same. Read Shannon.

An analogue data transmission has no discrete levels, any changes in amplitude vs time is the information it carries.
In contrast, a digital data transmission does have discrete levels, and variations in amplitude vs time, save for expected discrete levels, do not constitute as information.

You seem to mix up the carrier and the signal, and all your confusion stems from this. Signal is the intended information that is transmitted by the carrier. What information it is and how it's encoded and then extracted is up to the interpreter (sender and receiver) not up to the carrier.

 

[LightninBoy]

There is no such thing as digital transmission. You can transmit data that has been encoded into a digital format. But *ALL* transmission and *ALL* storage of digital data is analogue. The utility of digital encoding - the reason it was invented, the reason it has been so influential, the reason it has spawned an new age know as the Information Age - is that it enables transmission over analogue mediums with zero data loss and (as another poster correctly pointed out) storage and retrieval on analogue mediums with zero data loss.

Myself and several posters have tried to say this several times, but it is just being ignored by the OP.

 

[digitalfrost]

Well yes but that is getting philosophical. I could as well argue that in a world of rising entropy can we even have order and to not even bother because it's all just noise anyway.

 

[LightninBoy]

Well yes but that is getting philosophical. I could as well argue that in a world of rising entropy can we even have order and to not even bother because it's all just noise anyway.

Not philosophical to point out the difference between encoding and transmitting / storing. It is fundamental and real. Digital describes the encoding only. Once that encoding is represented in a physical form - which must happen for it to be stored or transmitted - that physical form is by definition analogue.

 

[JustJones]

Digital signal

To create a digital signal, an analog signal must be modulated with a control signal to produce it. The simplest modulation, a type of unipolar encoding, is simply to switch on and off a DC signal so that high voltages represent a '1' and low voltages are '0'.

 

[digitalfrost]

Not philosophical to point out the difference between encoding and transmitting / storing. It is fundamental and real. Digital describes the encoding only. Once that encoding is represented in a physical form - which must happen for it to be stored or transmitted - that physical form is by definition analogue.

I agree with that. But going back to the audiophile argument, and the outrage about MoFi using DSD for their vinyl releases where does this get us. Like you are technically absolutely correct. I come from a networking/computer background so I'm gonna say, on the physical layer obviously everything is analog because it has to be, but then on the data link layer we encode stuff digitally.

At the end of the day we live in an analog world. But like, that still doesn't make a digital signal analog.

 

[mansr]

You seem to mix up the carrier and the signal, and all your confusion stems from this. Signal is the intended information that is transmitted by the carrier. What information it is and how it's encoded and then extracted is up to the interpreter (sender and receiver) not up to the carrier.

That right there is all there is to this.

 

[kchap]

Any takers?

Not I. Whether you've paid for the five minute argument or the full half hour, I suspect you are unlikely to be satisfied with the answer.

 

[mansr]

Not I. Whether you've paid for the five minute argument or the full half hour, I suspect you are unlikely to be satisfied with the answer.

It'll just be contradiction anyway.

 

[RayDunzl]

Why is this so hard?

 

[pkane]

Why is this so hard?

It really isn’t.

 

[RayDunzl]

It really isn’t.

I wouldn't have thought so, either.