Is Digital Audio Transmission Analog? [video]

[Palladium]

Even a one bit error can be extremely audible - it just depends what bits you hit and hence how loud the "click" is.

PeteL seems to be confusing a couple of things - sure, skipping or repeating a single sample likely won't be terribly audible, because that doesn't introduce a large waveform discontinuity.

But actually changing the value of a single sample to something different from the surrounding waveform can be extremely audible as a click, depending on how much you changed it. You're potentially injecting a huge spike.

Digital systems are all designed to run with robust bit integrity in their entire transmission chain in mind and will fail spectacularly when it doesn't.

 

[pkane]

Isn't the scientific/technical side incredible enough ? Why do people seek to drag it back into the caves of the unenlightened ?

The modern cave dwellers simply don't want to leave. The shadows are too comfy, while sunlight is harsh and makes everything one knows about reality obsolete.

 

[PeteL]

Even a one bit error can be extremely audible - it just depends what bits you hit and hence how loud the "click" is.

PeteL seems to be confusing a couple of things - sure, skipping or repeating a single sample likely won't be terribly audible, because that doesn't introduce a large waveform discontinuity.

But actually changing the value of a single sample to something different from the surrounding waveform can be extremely audible as a click, depending on how much you changed it. You're potentially injecting a huge spike.

Sorry but a skipped sample is a waveform discontinuity, You can even pick one at the top of a wave at close to 0DBFS, the result will be equally inaudible. I don't get your point. UAC2 has ressources to detect errors, in early stages of development with things like XMOS environment, before you get things right, you may have a certain error rate. depending what it is it may click but a random error here and there will not, sorry.

 

[voodooless]

Well "In reality" transmission should be bit perfect. But extreme jitter problems will drop samples and it is actually what drop outs sounds like if there are insufficient buffering or clock rate mismatch.

Extreme jitter not necessarily leads to dropping of whole samples. It can also introduce bit errors. It totally depends on the source of interference.

Now, I don't agree that the fact that drop bits (not samples) are more audible because they are random.

The randomnesses is not the primary factor.

Why would that be? In a sample there is a "significance" grading. random means that bits with low significance are just as likely to be dropped than the most significant ones.

On average with a value of 0, the error vs the actual value is half as big as a totally random value. Also, more values tend to be around zero (nothing is always full volume).

Yes at least theoretically a most significant drop bit in a sample should be "more" audible" but if you can demonstrate than dropping a whole sample is not audible, it's not a far stretch to assess that a odd bit inside a sample will not be neither.

Well, errors generally do not occur as a singular instance. If that where the case, there is nothing to worry about.

 

[changster]

I find Darko's introduction in that video (with the water and then saying you show yourself out the door) very very insulting. He's degrading and insulting science.

Many people talk about how Amir is too negative and cuts into other people, but I find Darko's segment way way more insulting.

 

[PeteL]

Extreme jitter not necessarily leads to dropping of whole samples. It can also introduce bit errors. It totally depends on the source of interference.

Correct

The randomnesses is not the primary factor.

You did not respond to the question

On average with a value of 0, the error vs the actual value is half as big as a totally random value. Also, more values tend to be around zero (nothing is always full volume).

That's all very true but still don't get your point. A bit flip is a bit flip.. It is not automatically obvious, that's my sole point. You need a real problem to have click and pops, not a random error.

Well, errors generally do not occur as a singular instance. If that where the case, there is nothing to worry about.

Exactly my point.

 

[voodooless]

That's all very true but still don't get your point. A bit flip is a bit flip.. It is not automatically obvious, that's my sole point. You need a real problem to have click and pops, not a random error.

One single bit can already give you that if unclucky. It’s hard to notice obviously. You’ll need many errors for that, but I repeat myself

Exactly my point.

Good

 

[Vacceo]

I find Darko's introduction in that video (with the water and then saying you show yourself out the door) very very insulting. He's degrading and insulting science.

Many people talk about how Amir is too negative and cuts into other people, but I find Darko's segment way way more insulting.

Amir talks in a very cheerful way when he´s in an interview or when he explains stuff. If what he measures underperforms, well, that´s not negative, it´s underperforming.

The video where he talks with Joe Mariano is a perfect example of him speaking in a relaxed, nice and cheerful way.

The modern cave dwellers simply don't want to leave. The shadows are too comfy, while sunlight is harsh and makes everything one knows about reality obsolete.

I teach History of Art (among other things). I always tell my students that beauty, art, good flavors and many other delights in life get better the more you know about how they are created and work.

 

[anmpr1]

I really think you are selling Chinese fortune cookies short. Why do I say that? I remember awaiting news authorizing some important funding. For lunch the gang went to the Vegan Bodhisattva for an all you can eat buffet featuring Colonel Tso's Chicken D-Lite Stir Fry (latest scientific research has shown that chicken is actually a vegetable with feet).

And what do you know? Before the approval notice popped up in my email, I received authorization in the damned cookie! I was so impressed I kept the print out. Hook that up to your Audio Precision and run the spectrum, Mr. ASR!

That was the good news. The bad news was that about 2PM, I was hungry again.

 

[yuhoikuen]

A new argument has developed that transmission of digital audio is really analog. And for this reason, everything digital can be subject to audible difference from digital audio cables to digital output of streamers. This was emphasizes in a video by Darko Audio saying this explains his subjective opinion of streamers sounding different. I address this in this video and how there is some validity in what he says but his end conclusions are incorrect:

Schmitt trigger is used to overcome hysteresis in digital electronics. There is no such thing in analogue electronics. All these terms Jitter... are charaters of digital transmission. No matter how fast you drive a chip switching high and low, there is still hystersis and slew. Analogue transmission is a much simplier follower of the signal source, in transmission theory.
Analogue transmission does not neglect any part of the signal source waveform. It does not require approximation and reconstruction.

 

[lc6]

Let's do a test. While I'm not a fan of mqa - I avoid it as much as possible - mqa is great for doing this test.

Play an mqa through a 2 500 $ streamer, the mqa dac will display the logo.
Play it through a 100 $ streamer, the dac will still display the logo.

Both examples show the bits that were received are exactly the same, no matter if the analogue square wave was beautiful or not.

I'd invest in a good dac (mqa free in my case), dealing with jitter.
Darko probably wants me to rob a bank to get the more expensive streamer.

BTW, tests have shown that up to 1/3 of bits could be removed from an MQA file and it is still declared as "authentic." So much for the 'A' in the acronym.

 

[Tks]

This is the kind of speed you can get in consumer hardware. We can read 4 drives at the same time and combine the data without any faults. This is equal to streaming 136243(!) redbook cd's at the same time. Ok technically you wont be able to do that, because its not a sequential task, but you get the point.


View attachment 223373


Yeah, of course the LED is doing things and that needs to be interpreted at the other end but I was talking about the cable.

This is hilarious.

Now imagine someone takes a few seconds to set up something like a RAMDisk if they're going to get really picky about their storage speed, and watch theses figures skyrocket with higher-end RAM, especially the RND4K Q1T1 in both read and write.

Or if storage is a real problem for you, go with Sony's solution that has solved this issue once and for all with a dedicated product:

 

[American Pie]

BTW, tests have shown that up to 1/3 of bits could be removed from an MQA file and it is still declared as "authentic." So much for the 'A' in the acronym.

True, did the test myself. You can truncate a 24bit one to 16bits by leaving out every 3rd byte and it will still play, because mqa is actually pcm reduced to 15bit max + 1 bit signal + 8 bits lossy compressed inaudible freqs. (Still wondering why some prefer 15bits over 24bits btw).
Anyway you can't touch that one bit from the signal or the logo is gone.

 

[abdo123]

@amirm Your knowledge on signal processing is astounding and i'm glad you're sharing all of this with us thank you!

 

[Tks]

EDIT: Came to see a roasting, walked out a bunch more educated.

 

[poxymoron]

They do. On his video that I addressed here, there is a person saying he is going to buy that expensive streamer. And comment after comment saying this is one of his best videos explaining the technical aspects! He has far more subscribers than me currently so his voice carries farther.

These guys are up against the ropes, but they still have a lot of people in their corner.

 

[Ken Tajalli]

Amir claims it.
You are laughing at your host.
View attachment 223484

Please explain.
For the record, I respect Amir, but he is not my god, at times we may disagree, but I don't get your meaning in this case.

 

[abdo123]

@amirm I noticed that you mix your audio at a fairly low level to your video, usually i wouldn't bring these things up because at the end of the day it is your content but i really struggled to hear you with my laptop's speakers maxed out.

Looking at Youtube's audio normalization it looks like your audio RMS levels are around -25dBFS. if you don't compress your audio at all this is fairly low in my opinion.

 

[PeteL]

Please explain.
For the record, I respect Amir, but he is not my god, at times we may disagree, but I don't get your meaning in this case.

What is there to explain? You are saying that you laugh at the statement: "Is digital transmission analog?" and that whoever claims that has no conception of how digital work.
What I am saying is that Amir clearly state in the video that digital transmission is indeed analog. It was tongue in cheek, I don't think you are disrespecting Amir, but all I am saying is that he agree with this statement that you find laughable.

 

[Mulder]

They do. On his video that I addressed here, there is a person saying he is going to buy that expensive streamer. And comment after comment saying this is one of his best videos explaining the technical aspects! He has far more subscribers than me currently so his voice carries farther.