Is Digital Audio Transmission Analog? [video]

[dartinbout]

Ok grandad, 90%+ of my listening is electronic music, I find such genres as classical, rock, etc turgid and bloody awful to listen to.

Ignoring my first inclination to get trashy, as "turgid" implies something swollen, that could be interpreted as meaning electronica is somehow "flacid" but... yes I am older. Here's the catch, after dismissing most of the last 130 years of recorded music (and the the thousand years of composed music, preceding), I envy you your youth. You have a lifetime of discovery ahead of you. Although you may find it "bloody awful" now, age broadens taste. I hated opera, now I don't. I have recordings of Maria Callas that causes my flesh to tremble (no, I'm not having a stroke). I have live recordings of shows I saw, at the Fillmores', Winterland, Keystone Korner, Great America Music Hall and more. This 21st century offers so much auditory wealth. With the insane availability of so much music, at such ease, you will inevitably something new that is older then you, that you like. The other great thing is the equipment used to listen is so amazing (and cheap). You are really blessed with your lack of years, the future is bright (or at least in your old age, in an underground bunker, will ring with the sounds of "Daft Punk" to annoy your grandchildren).

 

[earlevel]

Increasing the number of bits not only reduces the quantization noise floor but also provides better time resolution (more bits can resolve smaller time differences).

That is true. But if it's dithered (or simply that the lowest bit is sufficiently below the noise floor, or the threshold of hearing), it becomes basically the same issue as time resolution with analog.

What I'm getting at is that if people don't care about time resolution in analog (they don't seem to—not talking about things like slew rate), we should probably drop that concern in digital too. Analog has that same issue with uncertainty in timing, due to noise floor or hearing—you can't be absolutely certain where the onset is.
To be clear, I', not correcting you, just using the opportunity to say this "out loud". If someone uses the "time resolution" argument as to why digital is inferior, I would say that it's no more of an issue than it is in analog.

 

[PeteL]

Increasing the number of bits not only reduces the quantization noise floor but also provides better time resolution (more bits can resolve smaller time differences).

Can you elaborate on "smaller time differences"?

 

[earlevel]

OK, Indeed I should let go of this semantic debate, it's not fully wrong to talk of resolution, all I wanted to clarify is that a waveform will be reproduced exactly the same way in 16 bits or in 24 bits. This waveform will only be closer to the noise floor in 16 bits. That's what I was not sure you where fully getting.

No problem. Looking back, it didn't help that when you referred to what the OP said, I take OP to be the original poster of the thread (but you meant someone I was replying to), so I wasn't getting your meaning fully.

Anyway, yeah, I'm pretty versed in finite word length effects, I'm been doing audio DSP since the '80s.

Cheers

 

[earlevel]

Can you elaborate on "smaller time differences"?

He means this:
Time resolution of digital audio—Troll Audio

I argue there another consideration that makes it even less of a property to digital in particular, making more or a issue with noise floor (therefore the same problem with analog), FWIW:

Time resolution in digital audio—EarLevel Engineering

It's a matter of viewpoint, but here I'm sort of doing as @PeteL argued (oh dam, that's who I'm replying to!), that we might as well think of the end product (we care about the timing as an auditory cue, right?), where the only thing that matter is the noise floor, as an influence for "timing resolution" in this context.

 

[PeteL]

He means this:
Time resolution of digital audio—Troll Audio

I argue there another consideration that makes it even less of a property to digital in particular, making more or a issue with noise floor (therefore the same problem with analog), FWIW:

Time resolution in digital audio—EarLevel Engineering

It's a matter of viewpoint, but here I'm sort of doing as @PeteL argued (oh dam, that's who I'm replying to!), that we might as well think of the end product (we care about the timing as an auditory cue, right?), where the only thing that matter is the noise floor, as an influence for "timing resolution" in this context.

Thanks

 

[DonH56]

Can you elaborate on "smaller time differences"?

The aperture time, or minimum time period that can be resolved, "time resolution" if you will, depends upon the signal frequency and the number of bits and not actually the sampling rate.

See here: https://www.audiosciencereview.com/forum/index.php?threads/digital-audio-jitter-fundamentals.1922/ -- the article is about jitter, but starts off with a picture and graph showing the effective time resolution vs. number of bits over frequency.

HTH - Don

 

[pseudoid]

You know how when you take the Flowflex COVID-19 Antigen Home Test and the little indicator thingy tells you "YES"? I wanted to shout out "Which COVID variety?" << B.1.1.7,? B.1.351? P.1? B.1.525? P.2? BA.X? (etc.)
Similarly, when somebody starts describing digital's virtues (or its vices); similar question comes to mind about "Which DIGITAL encoding/format?" << L/PCM? NRZ/i? Manchester? I2S? AES/EBU? (etc.)

 

[P_M]

This Time Resolution thing is getting weirder the more I search for it. Is this even an official/formal specification or even terminology recognized by any major manufacturer or even research institutions ?
Can anyone point me to any technical bulletins (like from TI/BB, AKM, Cirrus, Wolfson, etc..) or any scholarly articles from research institutions that specifically call it out ? Until then I am going to assume its troll business.....and I would like to encourage people to stop talking about it.

 

[MRC01]

This Time Resolution thing is getting weirder the more I search for it. Is this even an official/formal specification or even terminology ? ...

What exactly do you mean by "time resolution thing"? If you mean that identifying each wave zero crossing is critical for timing, and the amplitude resolution of each sample determines how close you can get to the zero crossing point, then this comes straight from signal processing math.

 

[DonH56]

This Time Resolution thing is getting weirder the more I search for it. Is this even an official/formal specification or even terminology ?
Can anyone point me to any technical bulletins (like from TI/BB, AKM, Wolfson, etc..) or any scholarly articles from research institutions that specifically call it out ? Until then I am going to assume its troll business.....

I learned it as aperture time. If you don't believe the article I wrote, most any company that makes data converters will have a white paper or app note about it. See ADI, Maxim, LTC, TI, etc. For the basis, read Shannon-Nyquist.

Here is a an ADI article after a quick search: https://www.analog.com/media/en/technical-documentation/application-notes/AN-501.pdf The concept is the same though this is targeting RF applications.

 

[JSmith]

This Time Resolution thing is getting weirder the more I search for it. Is this even an official/formal specification or even terminology recognized by any major manufacturer or even research institutions?

Time Resolution in Digital Audio - The Science of Sound

After having covered the more or less well known basics of digital audio in terms of bit depth and samplerate, I’d like to go one step further and explore another common buzzword: the time domain resolution of digital audio. If you have read more than one article here, it might have become...

science-of-sound.net

Time resolution in digital audio | EarLevel Engineering

www.earlevel.com

JSmith

 

[beeface]

ironically and quite obviously Johnny Darko is Australian, not Austrian

He's British, but spent time living in Australia.

Still, I'm curious what you mean by this comment?

 

[JSmith]

He's British, but spent time living in Australia.

Still, I'm curious what you mean by this comment?

Correct, in Sydney... and now resides in Berlin.

I also wondered what that poster meant by that comment.


JSmith

 

[Axo1989]

BTW, if you can stand the acid reflux of browsing Darko´s page (I hate to increase his traffic), there is a letter from Erin next to Amir´s. I will not give Erin crap; he may be more diplomatic, he may have a chrematistic interest and so on, but he does measure and he provides data, thus setting him in a different galaxy than the Darkos of this world.

It´s very interesting to see that, once boiled down, his message tells the same story as Amir´s yet the answer is completely different. ...

Of course the answer is different. Erin comes across as a guy who can present his knowledge and experience without making every disagreement a cage fight. Amir, unfortunately, opens with a pointed attack on "Karman" and concludes with a veiled threat, so Darko pegs him as a nutter with an axe to grind. Which is a shame, it could have been a good podcast session.

 

[Axo1989]

Correct, in Sydney... and now resides in Berlin.

I also wondered what that poster meant by that comment.

JSmith

As an Australian who went to high school in the US, being asked about Austria goes with the territory. It's a reference to the Austrian corporal of course per @Vacceo's Godwin-ism and follow-up.

 

[Axo1989]

I was wondering if there was a response to this video... then stumbled across that and thought, this is very poor form by Darko. To post a private message or email exchange is a shocker. ...

Ok, but it's a business communication, Amir is formally asking/offering to appear on a Darko podcast. Unless there is more to the exchange than we've seen, he didn't request confidentiality. You can make a distinction as to specific context perhaps but Amir usually favours publication in a business context and I generally agree (and did so in that thread) that such communication should be written carefully in the expectation that it is public:

No communication with a manufacturer from whom you bought a product is "private" information unless specifically agreed to. I have a number of times posted my frustrated chats with cable companies and such here. Nothing unethical about that. Companies need to know that anything they write in this context can be public information. Otherwise, we would be losing the side of the consumer. Members should be able to voice their frustrations with companies by proving them using such documentation.

Edit: that was Amir, this was me:

As I provide product/services to clients, I write my emails very carefully. Takes ages and is a pain in the arse really (makes them less friendly/personal too, which is kind of a shame) but it's essential.

 

[P_M]

What exactly do you mean by "time resolution thing"? If you mean that identifying each wave zero crossing is critical for timing, and the amplitude resolution of each sample determines how close you can get to the zero crossing point, then this comes straight from signal processing math.

I learned it as aperture time. If you don't believe the article I wrote, most any company that makes data converters will have a white paper or app note about it. See ADI, Maxim, LTC, TI, etc. For the basis, read Shannon-Nyquist.

Here is a an ADI article after a quick search: https://www.analog.com/media/en/technical-documentation/application-notes/AN-501.pdf The concept is the same though this is targeting RF applications.

So basically its plain old jitter. And how much jitter is acceptable at worst case amplitude and frequency.
Those tech bulletins dont say anything in regards to quantization (number of bits) or sample rate as having any relation to that.
does this "time resolution" offer any new insight that these big names in the business have not caught on to yet ?

 

[DonH56]

So basically its plain old jitter. And how much jitter is acceptable at worst case amplitude and frequency.
Those tech bulletins dont say anything in regards to quantization (number of bits) or sample rate as having any relation to that.
does this "time resolution" offer any new insight that these big names in the business have not caught on to yet ?

Jitter is error. The aperture time (time resolution) of an ideal data converter is a function of the number of bits and signal frequency. Jitter exceeding the minimum (lsb-level) aperture error reduces the SNR of the converter. The article I wrote begins by defining the time resolution, and the ADI app note presents the same information. They both include the equations that relate the timing aperture to the number of bits and show how SNR can be degraded by jitter on the order of and exceeding the aperture time of the device. (The ADI paper derives the error and introduces the number of bits in equation 6; my article states it in the text of the paragraph explaining Figure 1.) The basis for all of this is derived from the sampling theorem. Any "big name in the business" of designing data converters would understand this; it has been known for many decades.

As I said before, it is a function of the number of bits and signal frequency; the sampling rate does not enter into the equation.

I am not certain what you do understand and are angry about.

 

[MRC01]

So basically its plain old jitter. And how much jitter is acceptable at worst case amplitude and frequency.
Those tech bulletins dont say anything in regards to quantization (number of bits) or sample rate as having any relation to that.
does this "time resolution" offer any new insight that these big names in the business have not caught on to yet ?

As I understand it, it's standard sampling theory. It's not a new insight. Anyone who works in the field should understand it. Incidentally, it means the timing accuracy of standard 44-16 CD is much better than the incorrect value of 1/44100 or 22.68 microseconds that is commonly quoted by audiophiles who don't know any better. So perhaps it is "new insight" to them?

PS: the Troll audio link above has an intuitive explanation that's easy to understand. But he didn't invent that, he's just explaining it for folks who are curious.