[bfreedma]

In the video somewhere he said this doesent impact stored data like that exe you mention. Network streaming music is something that is happening live and if you turn of router/switch power it goes silent with no music heard. If you look at a downloaded picture from internet or once its uploaded on the page you can turn of internet and you still see the picture. With ethernet streaming music as source the timing is more critical and humans are more sensitive to hear that then we recently thought acording to Hans youtube clip. These voltages dont seem untouchable for music reproduction from streamed ethernet. Or otherwise at least something is happening that i cant explain other then all my posted theorys in this thread.

Jitter, grain, sibilance, phase distorsion, time smearing, rf/emi that kind or distorsion/noise.

Sorry to be blunt, but it's clear you don't understand how Ethernet and Ethernet switching works. Of course if you pull the power, traffic will eventually stop - no audiophile switch will change that. But Ethernet is not real time with no error correction. In fact, it's exactly the opposite. Data travels with CRC and is resent if the checksum fails. Data packets are also buffered, so in fact, is not real time.

It feels like you are scouring the internet, desperately searching for things you don't understand to fling against the wall in the hope that something sticks.

Citing Hans isn't helping. Hans has repeatedly demonstrated he has no knowledge of digital data transfer - he's just another audiophile discussing topics he doesn't have domain knowledge of.

 

[dazzerfong]

In the video somewhere he said this doesent impact stored data like that exe you mention. Network streaming music is something that is happening live and if you turn of router/switch power it goes silent with no music heard. If you look at a downloaded picture from internet or once its uploaded on the page you can turn of internet and you still see the picture. With ethernet streaming music as source the timing is more critical and humans are more sensitive to hear that then we recently thought acording to Hans youtube clip. These voltages dont seem untouchable for music reproduction from streamed ethernet. Or otherwise at least something is happening that i cant explain other then all my posted theorys in this thread.

Jitter, grain, sibilance, phase distorsion, time smearing, rf/emi that kind or distorsion/noise.

No it doesn't - there's a buffer. Ethernet does not work in real-time. No digital signals can work instantaneously since there's discrete time for information sent. In fact, for low-latency applications, that actually is a problem.

https://www.presonus.com/learn/technical-articles/Digital-Audio-Latency-Explained

Also, you're misapplying what the video is talking about completely, so that's a complete non-sequitur.

Citing Hans isn't helping. Hans has repeatedly demonstrated he has no knowledge of digital data transfer - he's just another audiophile discussing topics he doesn't have domain knowledge of.

Did you watch the video? Doesn't talk about transfer at all - he's assuming that the data has already arrived perfectly, and is talking about reconstruction.

I actually watched his other videos, and they seem to be fairly objective and accurate. Some of the titles are a bit out there, but the explanations themselves are quite rational. Was there anything of him that contradicts this?

 

[bfreedma]

No it doesn't - there's a buffer. Ethernet does not work in real-time. No digital signals can work instantaneously since there's discrete time for information sent. In fact, for low-latency applications, that actually is a problem.

https://www.presonus.com/learn/technical-articles/Digital-Audio-Latency-Explained

Also, you're misapplying what the video is talking about completely, so that's a complete non-sequitur.



Did you watch the video? Doesn't talk about transfer at all - he's assuming that the data has already arrived perfectly, and is talking about reconstruction.

I actually watched his other videos, and they seem to be fairly objective and accurate. Some of the titles are a bit out there, but the explanations themselves are quite rational. Was there anything of him that contradicts this?

Didn't watch that video but have seen enough video of Hans talking about digital transmission to understand his knowledge level (or lack thereof). His video on the SOTA ethernet switch is hilarious. He spouts audiophile "theory" as fact and then discusses measured differences so far below audibility that they can't possibly be material. His paeans to how much he loves the improved sound of an Ethernet switch were enough for me.

 

[gregorio]

[1] In the video somewhere he said this doesent impact stored data like that exe you mention.
[2] Network streaming music is something that is happening live and if you turn of router/switch power it goes silent with no music heard.
[2a] If you look at a downloaded picture from internet or once its uploaded on the page you can turn of internet and you still see the picture.
[3] With ethernet streaming music as source the timing is more critical and
[3a] humans are more sensitive to hear that then we recently thought acording to Hans youtube clip.
[4] These voltages dont seem untouchable for music reproduction from streamed ethernet.
[5] Or otherwise at least something is happening that i cant explain other then all my posted theorys in this thread.
[6] Jitter, grain, sibilance, phase distorsion, time smearing, rf/emi that kind or distorsion/noise.

1. Correct!!!

2. No, network streaming music is NOT live! You presumably have at least a 100 megabit and maybe a gigabit ethernet network. Even a 100 megabits per second network is about 70 times faster than the bits per second required by CD and about 10 times faster than lossless stereo 192/24. So, the data recieved from your ethernet network HAS TO BE STORED (in a buffer) and then fed from the buffer to the DAC chip at the appropriate speed (~1.4 megabits per sec in the case of lossless 44/16). As it HAS TO BE STORED and as point 1 is correct, then "this doesn't impact" the data!

3. That's because it's already been downloaded (or uploaded), if you turn the internet off before it's been downloaded you will NOT "still see the picture". Music and of course audio in general is constantly changing and so is downloaded and stored (in a buffer) in short bursts, rather than with say documents, where you have to wait until the whole document is downloaded before opening it. This is called "streaming" (rather than just downloading) and applies to any digital data that constantly changes, so mainly audio and movies/videos.

3. Again, NO, the timing of the ethernet network is not critical in the slightest. The timing of the data through the ethernet network is going to be wrong by a factor of about 70 times (or 700 times if you've got a gigabit ethernet network)! The timing of the data sent to the converter is critical but that's between your recieving data buffer and the point of conversion, not your ethernet network which is fixed at 100mbit per sec!
3a. Yes, humans are sensitive to digital data timing errors (called "jitter") in the case of music we're sensitive to jitter down to around 200 nano-seconds. This isn't recent though, this was first published in 1974 and confirmed by numerous subsequent scientific/controlled tests. Again though, this has nothing to do with an ethernet network, this is between the data recieving buffer in your DAC and the point of conversion.

4. "These voltages" must be untouchable by an ethernet network because an ethernet network does not carry "these voltages", it just zeros and ones!

5. Indeed, you are ABSOLUTELY RIGHT "something is happening" and the reason you "can't explain" is because you refuse to even consider the proven and obvious alternative, that your perception is not perfect! So instead, what you do is come out with "theories" that defy all the actual facts of how digital data and ethernet works and if your theories were true, there would be no digital data, digital devices or a "digital age" in the first place!

6. None of which can occur because as you correctly quoted, none of this can "impact stored data" and the data is stored AFTER it's passed through the ethernet network!

G

 

[TiJo]

deleted

 

[gregorio]

[1] Are you being daft on purpose?
[2] I actually watched his other videos, and they seem to be fairly objective and accurate. Some of the titles are a bit out there, but the explanations themselves are quite rational.

1. I think he is. In his defence though, as he refuses to accept the obvious, then he doesn't have any other choice. Whatever other theory or explanation anyone cares to come up with, sooner or later they run into the brick wall of contradicting basic facts and simple logic and therefore "being daft on purpose"!

2. He does seem to be a very personable man and on the face of it, his explanations do seem rational. Add these two attributes together and he seems entirely trustworthy and believable. Unfortunately though it's largely utter nonsense, although it's not obvious unless you happen to know or research the actual facts. He's quite open about a significant number of people pointing out that he's wrong but just repeatedly states the same two fallacies, that he hears a difference, so it must exist and he hears it because he has a better system than those who don't, both of which are completely false claims. He simply doesn't seem aware that the differences he's often talking about can't be reproduced by any system, so he obviously can't be hearing something that his system isn't even producing! The same nearly 30 year old audiophile fallacy that defies even simple logic.

G

 

[bluenight]

But Ethernet is not real time with no error correction. In fact, it's exactly the opposite. Data travels with CRC and is resent if the checksum fails. Data packets are also buffered, so in fact, is not real time.

Real time or not whatever the values(voltages/electrical signals) the buffer read is not perfect and always identical and maybe can be altered in everything in the signal path way to the read buffer that is then sent to dac at least as network audio streaming is concerned( read tellurium q aproach, phase distorsion). So if using better uniquely designed ethernet cables/switches for audio purpose with better signal integrity then generic ones it can easily impact sound for us to hear. Of course everything i say is only my theorys of why it can sound different that i picked up from or connected the dots from other more knowing that i linked to in the entire thread. Or everything has to do with better RF/EMI rejection or otherwise outside of our or mine edge of knowleadge.

Furthermore i dont believe this idea of perfect or identical digital when streaming sound/video is concerned. Just from my experience with my Tv whatcing i can see/hear every different device used for streaming for exe same netflix stream have different picture quality and sound. Be it mine chomecast ultra vs apple tv4k or the internal apps on my LG oled tv each device source has distinct differences in terms of sharpness of picture, colour reproduction and sound.

 

[bfreedma]

Real time or not whatever the values(voltages/electrical signals) the buffer read is not perfect and always identical and maybe can be altered in everything in the signal path way to the read buffer that is then sent to dac at least as network audio streaming is concerned( read tellurium q aproach, phase distorsion). So if using better uniquely designed ethernet cables/switches for audio purpose with better signal integrity then generic ones it can easily impact sound for us to hear. Of course everything i say is only my theorys of why it can sound different that i picked up from or connected the dots from other more knowing that i linked to in the entire thread. Or everything has to do with better RF/EMI rejection or otherwise outside of our or mine edge of knowleadge.

Furthermore i dont believe this idea of perfect or identical digital when streaming sound/video is concerned. Just from my experience with my Tv whatcing i can see/hear every different device used for streaming for exe same netflix stream have different picture quality and sound. Be it mine chomecast ultra vs apple tv4k or the internal apps on my LG oled tv each device source has distinct differences in terms of sharpness of picture, colour reproduction and sound.

It's been explained several times why how you believe this works and your theories are completely misaligned with the reality of digital data transmission.

Where I'm really lost is why you believe the digital transmission of audio/video somehow follows different standards than all other digital transmission. Data, sound, or audio, all of the data is encapsulated in specifically constructed packets with error correction protocol. How, within a packeted data set, would audio be identified on the wire, impacted by mysterious forces that don't impact the exact same packet construction holding non audio data, then altered. Not altered randomly, but changed in a consistent way that is represented by repeated artifacts. And that those artifacts are consistent within one brand of wire, but different in every other brand?

If Ethernet is so prone to signal alteration, how does the internet work? How do banks rely on digital transactions? How do nuclear power plants operate safely when the Ethernet connections to various controls and data collectors isn't reliable.

One thing I can state from first hand experience - data centers, banks, and nuclear plants use plain ole Ethernet cables. If there was any reality to your claims, the regulatory bodies would require "audiophile grade" Ethernet cables and switches. There's a rather obvious answer why they don't.

Edit - forgot to mention tellerium. As previously stated, prose from a vendor with a profit motive that lacks evidence to back up claims is not compelling.

 

[gregorio]

[1] Real time or not whatever the values(voltages/electrical signals) the buffer read is not perfect and always identical and maybe can be altered in everything in the signal path
[2] So if using better uniquely designed ethernet cables/switches for audio purpose with better signal integrity then generic ones it can easily impact sound for us to hear.
[3] Of course everything i say is only my theorys of why it can sound different that i picked up from or connected the dots from other more knowing that i linked to in the entire thread.
[4] Furthermore i dont believe this idea of perfect or identical digital ...

Great, so not happy with contradicting the facts, you're now even CONTRADICTING YOURSELF! Honestly, how low is it even possible to stoop?

1. The values in the buffer is digital data, zeros and ones. Are you now arguing that digital data isn't digital? The digital data in the buffer must be 100% perfect, otherwise it's rejected and is resent until it is. If this were not the case, all ethernet networks and all the digital systems that rely on them, like the internet for example, could not work. Are you claiming the internet doesn't work?

2. There is no "audio purpose", ethernet networks do NOT carry any audio, they are digital data networks and only carry digital data! How can you not understand this basic fact? And again, how does a cable "with better signal integrity" improve on the 100% perfect signal integrity of generic cables, what in your mind is better than 100% perfect? And lastly, you YOURSELF stated "In the video somewhere he said this doesent impact stored data.", so you're now even contradicting yourself!

3. That's not true, you're contradicting even those you've linked to!

4. Then you've answered my question: "simple logic dictates that either your statement is false or that the digital age does not exist" - You must believe the digital age doesn't exist! This begs the question, why are you typing a message on a device that doesn't exist and then posting it on an internet forum that doesn't exist? Honestly, how can this sound remotely logical to you?

G

 

[TiJo]

@bluenight

Actually that is something you understand wrong.
Let's assume you are right, that there is some noise in the streamed data, but no bit flips. Essentially this means, I can still correctly see any 1 and 0, but they might be slightly off. So now in that case what you could do is just read it, and store it again with some value add/subtracted that eliminates the noise. After that I can store this somewhere on my hard drive or even in RAM for any amount of time. Now later, I read it and send it to my DAC to get some analogue signal. There is no difference to non-streaming now.

(The thought experiment is based on the assumption of no bit-flips. In reality, bit flips can happen. There probability is super low with modern algorithms. Here someone calculated the probability: https://stats.stackexchange.com/que...asonable-assurance-of-absolute-data-integrity
To summarize, the chance, that a bit flip happens to you is close to 0 or in other words, will probably never happen in a lifetime of music streaming)

Obviously, a modern computer uses much more sophisticated methods. Many of these methods can be mathematically proven to be correct. Unless you or any cable manufacturer understands how the proofs work, there is no point in believing them. I would rather trust someone who has studied computer science

Hopefully, I was able to give you an more comprehensible view on digital data transmission.

 

[bluenight]

3a. Yes, humans are sensitive to digital data timing errors (called "jitter") in the case of music we're sensitive to jitter down to around 200 nano-seconds. This isn't recent though, this was first published in 1974 and confirmed by numerous subsequent scientific/controlled tests. Again though, this has nothing to do with an ethernet network, this is between the data recieving buffer in your DAC and the point of conversion.

4. "These voltages" must be untouchable by an ethernet network because an ethernet network does not carry "these voltages", it just zeros and ones!

3a How can it has nothing to do with the ethernet network when all buffered and read information comes from the source which is the ethernet work? And that information( voltages/electric signals) from network is sent to dac to work with.

4 In detail how does the sound information i hear travel as ones and zeros? What is these ones and zeros? Voltages/electrical signals of some sort i asume. How can they translate to the sound of a flute for exempel that i hear in my system?

From what i understand the adc measures/record sound as samples in time as voltages around twice 20khz in time that be a 16 bit/44.1 khz track. So it will be around 44100 samples in 1 sec right? And this is sent as zeroes and ones somehow in my ethernet network to my dac that exctract these numbers but in reverse way in voltages? This buffer to read the sound information from does only exist in my dac? Its not part of the ethernet plug i saw the tear down aq vodka cable had some chipset at the end? Or at the rj45 connection slot at my streamer that blinks gren yellow lights?

I still need more details of those zeros and ones.

 

[bfreedma]

3a How can it has nothing to do with the ethernet network when all buffered and read information comes from the source which is the ethernet work? And that information( voltages/electric signals) from network is sent to dac to work with.

4 In detail how does the sound information i hear travel as ones and zeros? What is these ones and zeros? Voltages/electrical signals of some sort i asume. How can they translate to the sound of a flute for exempel that i hear in my system?

From what i understand the adc measures/record sound as samples in time as voltages around twice 20khz in time that be a 16 bit/44.1 khz track. So it will be around 44100 samples in 1 sec right? And this is sent as zeroes and ones somehow in my ethernet network to my dac that exctract these numbers but in reverse way in voltages? This buffer to read the sound information from does only exist in my dac? Its not part of the ethernet plug i saw the tear down aq vodka cable had some chipset at the end? Or at the rj45 connection slot at my streamer that blinks gren yellow lights?

I still need more details of those zeros and ones.

there’s a lot of information available here. https://ieeexplore.ieee.org/browse/standards/get-program/page/series?id=68

When you get up to speed on how Ethernet works,I think this discussion will be much more productive.

 

[dazzerfong]

Didn't watch that video but have seen enough video of Hans talking about digital transmission to understand his knowledge level (or lack thereof). His video on the SOTA ethernet switch is hilarious. He spouts audiophile "theory" as fact and then discusses measured differences so far below audibility that they can't possibly be material. His paeans to how much he loves the improved sound of an Ethernet switch were enough for me.

Ah, I missed out on that one. Guess I dodged a bullet there.

there’s a lot of information available here. https://ieeexplore.ieee.org/browse/standards/get-program/page/series?id=68

When you get up to speed on how Ethernet works,I think this discussion will be much more productive.

Let's be realistic, he ain't gonna read it. He won't have the acumen or time to pour through the standards. Not to mention, you gotta pay for it or be an IEEE member.

This is a waste of time - this is like an anti-vaxxer who, after reading 1 article on vaccination, thinks they know better than doctors.

1. I think he is. In his defence though, as he refuses to accept the obvious, then he doesn't have any other choice. Whatever other theory or explanation anyone cares to come up with, sooner or later they run into the brick wall of contradicting basic facts and simple logic and therefore "being daft on purpose"!

2. He does seem to be a very personable man and on the face of it, his explanations do seem rational. Add these two attributes together and he seems entirely trustworthy and believable. Unfortunately though it's largely utter nonsense, although it's not obvious unless you happen to know or research the actual facts. He's quite open about a significant number of people pointing out that he's wrong but just repeatedly states the same two fallacies, that he hears a difference, so it must exist and he hears it because he has a better system than those who don't, both of which are completely false claims. He simply doesn't seem aware that the differences he's often talking about can't be reproduced by any system, so he obviously can't be hearing something that his system isn't even producing! The same nearly 30 year old audiophile fallacy that defies even simple logic.

G

I mostly ignored his audibility discussion and focused on how he explained things. Guess that's where things get controversial: his technical explanation is sound, but the audibility aspects leave a bit to be inspired.

 

[JackFX]

No, they are built to a standard it is a protocol. Going from Cat5e to Cat6 or Cat7 is technically an upgrade and can run a further distance but the cables in your walls are likely of Cat5e or lower and unless you are running a really far distance not worth it.

 

[gregorio]

3a How can it has nothing to do with the ethernet network when all buffered and read information comes from the source which is the ethernet work? And that information( voltages/electric signals) from network is sent to dac to work with.
[4] In detail how does the sound information i hear travel as ones and zeros? What is these ones and zeros? Voltages/electrical signals of some sort i asume. How can they translate to the sound of a flute for exempel that i hear in my system?
From what i understand the adc measures/record sound as samples in time as voltages around twice 20khz in time that be a 16 bit/44.1 khz track. So it will be around 44100 samples in 1 sec right? And this is sent as zeroes and ones somehow in my ethernet network to my dac that exctract these numbers but in reverse way in voltages? This buffer to read the sound information from does only exist in my dac? Its not part of the ethernet plug i saw the tear down aq vodka cable had some chipset at the end? Or at the rj45 connection slot at my streamer that blinks gren yellow lights?

3a. It has nothing to do with the ethernet nertwork because all the data received is stored in a buffer AFTER the ethernet network and as you yourself have quoted, stored data is NOT impacted by noise, jitter or anything else! In other words, the "source" is effectively this buffer NOT the ethernet network.

4. You are asking: "What is binary digital data and how is it used to store and transfer analogue audio signals?" These are good questions but the OBVIOUS (and frankly incomprehensible) problem is that you should have asked and answered them BEFORE you started publicly arguing about it! How can you spend days and days publicly arguing about digital audio data when you don't even know what digital audio data is?!

The Sound Science sub-forum is probably the correct place to ask and answer your questions and answering "in detail" is not practical here, because there are a lot of details, which is why we need micro-processor devices (such as ADCs and DACs) that can deal with lots of details. So, very simplified (not "in detail") answers:

"What is these ones and zeros? - These ones and zeros are just that, the numerical values of "one" and "zero".

"Voltages/electrical signals of some sort i asume." - No, these ones and zeros ("bits") can be pretty much anything that has two states which can quickly and reliably be read, the most common examples: The ones and zeros can be regions of reversed and non-reversed magnetic fields on a tape, floppy disk or hard disk drive. They can be microscopic indentations ("pits") and non-indentions ("lands") on CD, DVD and BluRay disks (read by measuring the different reflections produced by a laser). They can be a sequence of on/off pulses of light in fibre-optic networks. They can be a high or low voltage stored in microscopic capacitors on SSDs, RAM modules, etc., or, they can be a sequence of high and low voltages in the case of ethernet, USB and similar networks. The important thing is that there are ONLY TWO states, it doesn't matter how distorted a reversed/non-reversed magnetic field is, how imperfect the "pits" and "lands", how degraded the on/off pulses of light or how noisy a high or low voltage is, as long as the two different states can be identified. A massively degraded/distorted pulse of light is still a pulse of light (as opposed to no light) and is therefore a "1". A massively distorted high voltage is still a high voltage (a "1") and a massively distorted low voltage is still a low voltage (a "0"). So, a perfect sequence of zeroes and ones can be stored and transferred even using highly imperfect electrical, optic or magnetic signals and media, which is precisely why binary digital data was invented in the first place!

"How can they translate to the sound of a flute for exempel that i hear in my system?" - On their own, a one or a zero is pretty useless but by combining them we can represent a far greater range of values. This is exactly the same as our decimal number system: We only have the numerical values of 0 - 9, which is a very small range of values but by combining them we can represent a very large range values. For a value beyond the number 9 we combine two numbers to form "10", beyond 99 we combine three numbers and so on. The exact same happens with binary, we combine our numbers (bits) to represent larger values: We have 0 and 1, then we need another bit for the next number, 10 (= our number 2), the next number 11 (= 3) is still two bits but the next number is 100 (= 4) which requires 3 bits, 101 = our number 5 and so on. CD uses a combination of 16 bits (16 zeros or ones), which can represent any value up to 65,536.

What an ADC does is measure the voltage of an incoming analogue audio signal and then, using a mathematical formula, assigns it a value between 0 and 65,536 (represented by a sequence of 16 ones and zeros). In the case of CD, 44,100 measurements and digital value assignments are made per second. A DAC takes these digital values (sequences of 16 zeros and ones) and using a mathematical formula, reconstructs the original analogue audio signal. The answer to your question is therefore: They CANNOT translate to the sound of a flute! The ONLY thing that ADCs and DACs "translate" is an analogue audio signal into a sequence of zeros/ones and back again, so it can be stored and transferred perfectly. That's it, there is NOTHING else, there is NO sound! Sound doesn't exist until the very end of the reproduction chain, which is why you need headphones or speakers, to convert the analogue audio signal into sound. Whether the sound you hear sounds like a flute, depends on the analogue signal, how well it's converted to sound (by your HPs) and your brain's interpretation (perception) of that sound.

G