Is Digital Audio Transmission Analog? [video]

[danadam]

It's a waveform, it has a frequency, it has an harmonic content, it is subject to noise and distortion, it can be represented on a FFT, just out of the audio band. it has all the attributes of an analog signal. What you do with it afterward is not what define the signal itself.

I guess I have a more abstract definition of what a signal is.

I guess that yes we could technically use the term "signal" by what is actual information, the content delivered, then yes digital would be appropriate but it's a bit far fetch.

Really? Seems natural to me.

 

[danadam]

Analog implies something that is analogous. The rise and fall of the voltage level is analogous to the compression and rarefaction pattern of sound in the air. Clearly, it pushes a speaker to create those exact compression and rarefaction patterns.

In the most strict meaning of the word, yes. But from All amplifiers are analogue, but some amplifiers are more analogue than others (AES Convention Paper 353, B. Putzeys, A. Veltman, P. van der Hulst and R. Groenenberg):

The
meaning of the word analogue is generally extended
to apply to the original physical quantities as well,
so the air pressure itself is also called analogue. In
other words, physical quantities such as voltage, cur-
rent, position, pressure and time are all analogue.

And another quote I like from that paper about what a digital signal is:

What makes a signal digital is whether the recipient
interprets it as such i.e. if there has been an agreed
method of coding symbols into it, used by both ends.
In a “digital signal”, analogue is the form, digital the
content. A square wave can be digital or analogue,
a continuous-time signal can be digital or analogue
too. It all depends on what happens downstream.

 

[PeteL]

I guess I have a more abstract definition of what a signal is.


Really? Seems natural to me.

I guess we can play with words all day and night. Bottom line what goes trough the wire is an oscillating voltage subject to perturbations and physical limitations. It's not data. For the rest I agree with the statement, Is digital transmission analog? Yes. I should just not have brought up the word signal and stick to this wording.

 

[danadam]

One single bit can already give you that if unclucky.

I made some bit-flip examples on another forum.

 

[MRC01]

Given the great amount of trigonometrical/periodic functions that arise in signal processing, I would be so confident using rational numbers in this context.

I just commented on that because nobody uses rational numbers to represent a physical magnitude. Some of those gaps are quite meaningful and arise quite often, like sqrt(2) or pi.

Mathematically speaking, sure.
Pragmatically speaking, for any real number R, for any epsilon > 0, you can find a rational number Q such that | R - Q | < epsilon.
You can pick epsilon smaller than any means of measurement or detection.

 

[diaolodoro]

pma said:
“Digital” is a mathematical representation. Signal transfer is through voltage and voltage is an analog quantity. Digital signal representation is transferred by means of analog quantity - voltage.

No it ain't.
Look up definition for analog .
A series of on/off pulses is not analog.
it only has two states.
Analog has undefined states.

Digital is nothing more than a useful technique, or like pma said, a mathematical representation. Even the data storage on our computers is just different levels of stored electrons on memory cells. Theoretically everything in life is analog. Isn't it beautiful?

 

[P_M]

The title of the video is probably deliberately and artificially made sensational, but the subject is explained a bit more clearly in the first post and image here -

The Sound of a Digital Cable: Bandwidth and Jitter

Can your digital cable impact the sound of your system? Like most things, the answer is “maybe”… Reading some of the cable advertisements I see all kinds of claims. Some make no sense to me, such as “use of non-ferrous materials isolates your cable from the deleterious effects of the Earth’s...

www.whatsbestforum.com

Basically, digital (square) waveform needs infinite (or at least very high) bandwidth to transmit like a square otherwise it comes out wonky and wrecks havoc for the digital waveform interpretation. In audio it causes jitter and also interpolated (instead of actual) data.
If you look at a FFT of a 1Khz square wave you will see that its littered with 20+ harmonics. These harmonics are what make the difference between a 1Khz sine wave and 1Khz square wave. And all those harmonics including that 20th+ one have to be transmitted properly to reveal the same square wave on the receiving end. So if you want a 1Mhz square wave to look like a 1Mhz square wave on the other end of the wire (optical or coax or ethernet or even wireless) you want at least a flat 20Mhz bandwidth of the entire transmission layer (transmitter, receiver, cable etc..). because digital systems rely on the transition edge of that square wave to determine 1 or 0 or other control like for example clock ticks.
Otherwise you loose those higher harmonics and that results into not square and straight edges but bent corners and edges which if bent far enough can cause that interpretation of 1 and 0 and clock ticks to be way off than the original. This causes bandwidth-limited transmission jitter on spdif and similar transmissions methods.

http://nwavguy.blogspot.com/2011/02/jitter-does-it-matter.html (scroll down to cable jitter section and see the image for it).

edit: and btw, before usb asynchronous came out it was usb isochronous and probably some devices still use it, but it was a complete disaster in terms of jitter due to this reason. because loss of data in data transmission over usb or ethernet can be detected and resolved on the fly but there is no room for such correction for real-time audio data. That is why it works great for transmitting GoBs of data but not realtime audio, and worse, from which a clock needs to be recovered.

and here's a nice animation demonstrating the high number of harmonics that need to be passed in order to preserve a good square form -

https://i0.wp.com/wtt.pauken.org/wp-content/uploads/2011/04/Synthesis_square.gif?ssl=1

 

[earlevel]

Close but I would quibble with the definition of digital.

Digital is about representing information as discrete states, nothing in between the states. Can be binary, trinary, or whatever. Not necessarily numbers. In fact, numbers can be a continuum (rational numbers, real numbers). Digital states may be represented as discrete numbers, but they don't have to be numbers, and numbers are incidental. The essential point is that the states are discrete.

I disagree that "digital is about discrete steps". There are no advantages to discrete steps, it's just an unavoidable, practical consequence of being about numbers. A good hint of that is that we strive to reduce the step size as much as practical (24-bit audio, 32- and 64-bit float processing). Sampling is about discrete time, digital is about storing them as numbers. Discrete steps are about neither, just about resolution.

"...they don't have to be numbers, and numbers are incidental."—Oh? As a DSP guy, I find that pretty statement pretty hard to defend—care to try? I don't care if you want to call them symbols or whatever, in the end you've only made up a new numbering system. How are you going to do a gain change if they aren't numbers? Filter? FFT? You'll need to do math, with numbers.

Analog implies a smooth continuum of signal. Not discrete, but any of infinitely many values within a range is valid. Note that this does not imply that analog has infinite resolution. Some states are so close together they cannot be differentiated.

You said analog means a representation that is "analogous" to the information being represented. This is true. Yet all our senses are perceptually "analog" -- regardless of whether the universe is actually analog or digital, smooth continuous is how we perceive it. So as a consequence analog ends up implying smooth continuous.

This just comes down to what words you like to use, but we're talking about audio in this thread, not the sense of touch or taste. As such, I don't think it does anyone any good to call something like AES-EBU "analog". It's a digital signal in the continuous time domain—more simply, "digital". Analog may imply "a smooth continuum of signal", but a continuous signal doesn't necessarily imply an analog signal.

It won't make me mad if you want to call AES-EBU "analog", I'm just saying I don't think it helps anyone. We all know what it means when someone says "analog audio". It's something that, on an oscilloscope, looks like what we expect the speaker to do when it's playing back that audio.

 

[pablolie]

I don't know what's so hard about this.

The *ADVANTAGE* of digital is that it ensures signal integrity. Throughout *any* transport media that meets basic specs.

There's something called the OSI layer standard that walks people through this [1]. Physical signal transmission is always analog. Its interpretation is not. Look at stuff like PAM4 for the next gen of ultimate speed. The *challenge* is to find the analog system to deliver a flawless, error-free digital signal. One that is good enough for mission-critical apps way more important than a HiFi audio source. Hence... trust you get error free digital audio every time unless you have a really questionable setup..

[1] It basically means that the layers work together in delivering on the requested level of service. And it works perfectly for applications way more critical than home audio.

PS: To provide a metaphor... think about calling a friend live from a loud venue. Analog is ... you can't because nobody understands. . Digital is... you text and the distortion doesn't matter - your message is delivered error-free. On top of the exact same transport.

 

[danadam]

edit: and btw, before usb asynchronous came out it was usb isochronous

It is still an isochronous transfer. Isochronous transfer is used for real-time data and has 3 modes: synchronous, adaptive and asynchronous. I've never seen synchronous DAC, I know ODAC was adaptive and the rest I have (and probably every one produced since some time) are asynchronous.

Beside isochronous transfer there are: bulk (for data), interrupt and control.

 

[DonH56]

The title of the video is probably deliberately and artificially made sensational, but the subject is explained a bit more clearly in the first post and image here -

The Sound of a Digital Cable: Bandwidth and Jitter

Can your digital cable impact the sound of your system? Like most things, the answer is “maybe”… Reading some of the cable advertisements I see all kinds of claims. Some make no sense to me, such as “use of non-ferrous materials isolates your cable from the deleterious effects of the Earth’s...

www.whatsbestforum.com

Basically, digital (square) waveform needs infinite (or at least very high) bandwidth to transmit like a square otherwise it comes out wonky and wrecks havoc for the digital waveform interpretation. In audio it causes jitter and also interpolated (instead of actual) data.
If you look at a FFT of a 1Khz square wave you will see that its littered with 20+ harmonics. These harmonics are what make the difference between a 1Khz sine wave and 1Khz square wave. And all those harmonics including that 20th+ one have to be transmitted properly to reveal the same square wave on the receiving end. So if you want a 1Mhz square wave to look like a 1Mhz square wave on the other end of the wire (optical or coax or ethernet or even wireless) you want at least a flat 20Mhz bandwidth of the entire transmission layer (transmitter, receiver, cable etc..). because digital systems rely on the transition edge of that square wave to determine 1 or 0 or other control like for example clock ticks.
Otherwise you loose those higher harmonics and that results into not square and straight edges but bent corners and edges which if bent far enough can cause that interpretation of 1 and 0 and clock ticks to be way off than the original. This causes bandwidth-limited transmission jitter on spdif and similar transmissions methods.

http://nwavguy.blogspot.com/2011/02/jitter-does-it-matter.html (scroll down to cable jitter section and see the image for it).

edit: and btw, before usb asynchronous came out it was usb isochronous and probably some devices still use it, but it was a complete disaster in terms of jitter due to this reason. because loss of data in data transmission over usb or ethernet can be detected and resolved on the fly but there is no room for such correction for real-time audio data. That is why it works great for transmitting GoBs of data but not realtime audio, and worse, from which a clock needs to be recovered.

and here's a nice animation demonstrating the high number of harmonics that need to be passed in order to preserve a good square form -

https://i0.wp.com/wtt.pauken.org/wp-content/uploads/2011/04/Synthesis_square.gif?ssl=1

Thanks! First article is reproduced here: https://www.audiosciencereview.com/...ital-cable-bandwidth-with-eye-diagrams.20539/ The original is over ten years old; now most if not all DACs use clock recovery circuitry that isolate them from the incoming data stream.

And there is this on square waves: https://www.audiosciencereview.com/.../composition-of-a-square-wave-important.1921/

Several articles on jitter etc. in the post linked in my sig.

 

[nagster]

I remember being taught in school that an Elementary charge is the minimum unit of electricity.
Is there a minimum unit for voltage or current?

 

[DonH56]

I remember being taught in school that an Elementary charge is the minimum unit of electricity.
Is there a minimum unit for voltage or current?

What I was taught was one electron (or hole).

 

[PeteL]

I've never seen synchronous DAC,

Any DAC when SPDIF or AES-EBU is used?

 

[danadam]

Any DAC when SPDIF or AES-EBU is used?

I meant USB, isochronous transfer in synchronous mode.

 

[DonR]

I remember being taught in school that an Elementary charge is the minimum unit of electricity.
Is there a minimum unit for voltage or current?

current is charge carriers over time so since time can be infinite, there is no smallest measure. If we choose 1 second of time (standard for amperes) then 160 zeptoamperes (roughly) is 1 electron per second.

 

[P_M]

It is still an isochronous transfer. Isochronous transfer is used for real-time data and has 3 modes: synchronous, adaptive and asynchronous. I've never seen synchronous DAC, I know ODAC was adaptive and the rest I have (and probably every one produced since some time) are asynchronous.

Beside isochronous transfer there are: bulk (for data), interrupt and control.

yep mixed up the terminology there. It has been years probably over a decade since I played around with that stuff so a bit rusty. I do remember though in the early days of usb audio, synchronous dacs were the only types of dacs and were hated - probably not even dacs per se, in that era there were just "external soundcards".

 

[Daverz]

Given the great amount of trigonometrical/periodic functions that arise in signal processing, I would be so confident using rational numbers in this context.

I just commented on that because nobody uses rational numbers to represent a physical magnitude. Some of those gaps are quite meaningful and arise quite often, like sqrt(2) or pi.

 

[P_M]

Thanks! First article is reproduced here: https://www.audiosciencereview.com/...ital-cable-bandwidth-with-eye-diagrams.20539/ The original is over ten years old; now most if not all DACs use clock recovery circuitry that isolate them from the incoming data stream.

And there is this on square waves: https://www.audiosciencereview.com/.../composition-of-a-square-wave-important.1921/

Several articles on jitter etc. in the post linked in my sig.

Didn't know you were here
Thanks.

 

[amirm]

Recently listened to Darko‘s podcast episode with golden sound. If you want to understand darko‘s opinion on measurements, I recommend listening to it.
I generally don’t recommend trying to argue with tryhard subjectivists because it’s mostly just a waste of time.
But Darko is a bit more open towards different opinions as the podcast has shown IMHO

While he does come across that way on the podcast, that is not his real position. To wit, I followed up with him on that very thing and offered to come on his show and explain these things. He turned me down.