How COULD upgraded Ethernet cables make a positive difference? What's behind it?

[amirm]

If by data you mean measurements then no, it's too early - this is still at the investigative stage & like any such investigation evidence of effect & circumstances are being accumulated.

Thanks John. To be very clear, I am still asking about differences in Ethernet cables. If this is what is being investigated, do you mind saying more? Who is doing the investigation? What is the hypothesis they are pursuing? What work has been done so far?

This may eventually lead to some solid measurements but it will only come from people who understand the underlying premise for what is going on & as a result, employ careful measurement techniques designed to uncover the mechanism - it won't come from those who employ standard measurements & don't understand the noise that is being discussed here.

Well, this is why I asked the question above with respect to who is doing this work. The level of system knowledge required for networked streaming goes far beyond DIY hardware hobbyist who bangs together a DAC or USB interface. There is a lot of false assumptions being made about how systems works as I explained in this post regarding how Ethernet streaming works. What wireworld is saying there is just flat out wrong. But I don't know everything and hence my question as to who is doing investigation.

By the way, I was not necessarily asking for measurements. You mentioned there would be audible effect so I was interested in any listening tests that had shown that. Is there any such data?

I have no direct experience with ethernet cables making an audible difference but have with USB cables & USB transmission - that's my data point. Not too dissimilar operational factors, differential, packet based protocol but ethernet is pure differential & can make use of a transformer, conferring a certain amount of isolation while USB needs to drop back to single ended transmission for certain messages & therefore can't use transformers.

There is a heap of software above the physical network layer in Ethernet that has no resemblance whatsoever to what happens in USB. Without understanding that, the person would have no prayer of knowing what is going on or how to test for it. Systems using Ethernet for example routinely run Linux OS but no such OS is needed to make a USB interface for example.

 

[jkeny]

Thanks John. To be very clear, I am still asking about differences in Ethernet cables. If this is what is being investigated, do you mind saying more? Who is doing the investigation? What is the hypothesis they are pursuing? What work has been done so far?

Well before you jump to the question about differences in cables, you need to have a premise about what possible mechanism could result in audible differences. The mechanism being premised here & elsewhere is that certain types of noise can be transmitted between devices in an audio replay system that audibly effect it's output. This really is not news - it's a well known phenomena & something that AES has produced a standard for AES48-2005. Now what is the scope of this standards document

"This standard specifies requirements for the termination, within audio equipment, of the shields of cables
supporting interconnections with other equipment, taking into account measures commonly necessary for
the preservation of EMC (electromagnetic compatibility) at both audio and radio frequencies"​

Does this document concern itself with the audible effects of such EMC problems? No!

I consider what is going on in this area in both USB & ethernet connections in audio to be a furtherance of this known problem but with an audibility focus.

Getting back to cabling - can I envisage cabling differences between cables, all of which work fine for digital data transmission, having differences in the amount & frequency distribution of differential & common mode noise? Yes, absolutely! Will this translate into audible differences? Not always.

Well, this is why I asked the question above with respect to who is doing this work. The level of system knowledge required for networked streaming goes far beyond DIY hardware hobbyist who bangs together a DAC or USB interface. There is a lot of false assumptions being made about how systems works as I explained in this post regarding how Ethernet streaming works. What wireworld is saying there is just flat out wrong. But I don't know everything and hence my question as to who is doing investigation.

Yes, I agree with you - data errors are not the underlying cause.

By the way, I was not necessarily asking for measurements. You mentioned there would be audible effect so I was interested in any listening tests that had shown that. Is there any such data?

Private listening data, yes, absolutely!

There is a heap of software above the physical network layer in Ethernet that has no resemblance whatsoever to what happens in USB. Without understanding that, the person would have no prayer of knowing what is going on or how to test for it. Systems using Ethernet for example routinely run Linux OS but no such OS is needed to make a USB interface for example.

As I said, this is not a software issue so I don't know why you want to go there? You seem determined to avoid what has been told to you many times that differential & or CM noise are likely at the heart of this - you can't keep building strawmen to knock down .

 

[amirm]

John, I don't know what that answer means. You said there was investigation going on. I am asking who is doing the investigation and the status of it. Does this question have an answer?

 

[YashN]

Yes, I agree with you - data errors are not the underlying cause.

Are you sure this is what Amir thinks? Because he writes the same erroneous assertion that it is about the data here:

In such a short cable, you have to really, really go out of your way to corrupt any data.

 

[jkeny]

John, I don't know what that answer means. You said there was investigation going on. I am asking who is doing the investigation and the status of it. Does this question have an answer?

Ah well, I did try!

I already stated where this research is at but it appears that you don't seem to understand? There are various stages in any research - the first of which is a broad question for research - such as "can different ethernet cables audibly affect sound?" which needs to be narrowed down & honed into a testable hypothesis - mostly through observation & fitment of these observations to known mechanisms - often extending these mechanisms.

At least one possible hypothesis has been formulated but testing of it requires some specific skills & measurement techniques.

However, with or without measurements we always need to personally evaluate with listening so in essence you want measurements but this is not a prerequisite for audibility - in fact I would say they are relatively orthogonal.

 

[jkeny]

Are you sure this is what Amir thinks? Because he writes the same erroneous assertion that it is about the data here:

I don't know - I was looking for some sort of common ground but he throws up so many strawmen that it's difficult to know what he really thinks - his arguments remind me of this classic

 

[amirm]

Ah well, I did try!

It might just be me but I am still trying to find out who is doing this work:

If by data you mean measurements then no, it's too early - this is still at the investigative stage & like any such investigation evidence of effect & circumstances are being accumulated.

Who is doing the investigation John? Where is that data that has been accumulated already?

 

[jkeny]

It might just be me but I am still trying to find out who is doing this work:


Who is doing the investigation John? Where is that data that has been accumulated already?

http://www.whatsbestforum.com/showt...at-s-behind-it&p=424729&viewfull=1#post424729

 

[jkeny]

 

[NorthStar]

? http://www.howtogeek.com/210326/not...l-you-can-get-faster-lan-speeds-by-upgrading/

"Most people won’t really care whether they’re using Cat-5e or Cat-6 cables at home. The Internet connection is the bottleneck, and Cat-6 cables won’t help that. Cat-6 can enable faster speeds when transferring files or otherwise communicating between two computers on the local network, but most people won’t notice.

Still, there is a difference! If you’re wiring your home with cables that will be stuck there a while, you should definitely go for Cat-6 for the future-proofing and faster LAN speeds.."
______

• http://www.howtogeek.com/70494/what-kind-of-ethernet-cat-5e6a-cable-should-i-use/

 

[amirm]

Are you sure this is what Amir thinks? Because he writes the same erroneous assertion that it is about the data here:

I am happy to be corrected that a 6 foot Ethernet patch cable generates errors. If that were so, and the cables in question here did anything to improve that situation there would be a massive market for them in computer/data center for world.

As I mentioned in my ASR Forum post, measuring data errors on our home networks is trivial. Anyone who thinks there are such errors should first go and get the statistics to see if that is the case.

To that end, I just went and pulled the stats from my dlink "smart" (web-enabled) 48 port gigabit switch that runs my entire home (more or less). I pulled up the per-port stats for the last 58 days it has been up and there is not a single transmit or receive error and this is over long pulls of ethernet cabling throughout my home. Here is an example snapshot:

Take a look at port 44 that I have highlighted: 761,309,868 packets (chunks) of data has been transmitted with no errors. Likewise, that port has received 3,833,536,602 or nearly four billion packets with absolutely no errors!

You can buy 8 port smart switches for the price of week's worth of starbuck's coffee . Just get one and test such hypothesis. If you do find errors, I love to see them.

Until then, this dog don't hunt. Have Mike ask his son if he would use the wireworld ethernet cable to reduce errors in his job. I am confident his answer would be no way.

 

[cjf]

I pulled up the per-port stats for the last 58 days it has been up and there is not a single transmit or receive error and this is over long pulls of ethernet cabling throughout my home. Here is an example snapshot:

Take a look at port 44 that I have highlighted: 761,309,868 packets (chunks) of data has been transmitted with no errors. Likewise, that port has received 3,833,536,602 or nearly four billion packets with absolutely no errors!

I would like to see the same Stats from a Wire World or Audioquest ETH Cable user but more interested in the WW cable since its geometry goes against the grain of what a "normal" ETH cable looks like under the wrapper. If I were a believer in esoteric ETH Cables I would buy one myself and do the test but since I have my doubts on what its contribution would be, other than being more likely to generate Errors/Discards, I'll have to wait for some other brave soul to tread those waters

 

[jkeny]

I would like to see the same Stats from a Wire World or Audioquest ETH Cable user but more interested in the WW cable since its geometry goes against the grain of what a "normal" ETH cable looks like under the wrapper. If I were a believer in esoteric ETH Cables I would buy one myself and do the test but since I have my doubts on what its contribution would be, other than being more likely to generate Errors/Discards, I'll have to wait for some other brave soul to tread those waters

Can you explain how you fall for this strawman argument about data errors, thereby furthering it & yet you use this in your rig BlackBox - Ethernet Opto-Isolator ?

 

[amirm]

Can you explain how you fall for this strawman argument about data errors, thereby furthering it & yet you use this in your rig BlackBox - Ethernet Opto-Isolator ?

Strawman? No, I am going by specific statements made by wire world about their Ethernet cable:

Category 7 cabling was created to satisfy the demands of 10 Gigabit Ethernet. According to Wireworld, “Even though most media networks now run below that speed, cables that support higher speeds have been found to improve the quality of audio and video streaming. Those improvements are possible because streamed signals suffer from data errors that cannot be repaired by the error correction systems that preserve file transfers."

As I have explained and showed, this is complete falsehood. Data errors are hugely unlikely to occur in your home and upper level layers of software most definitely perform error correction. It is pure false advertising.

 

[cjf]

Can you explain how you fall for this strawman argument about data errors, thereby furthering it & yet you use this in your rig BlackBox - Ethernet Opto-Isolator ?

I fail to see how the use of an Opto Isolator applies here. At least the Opto Isolator does serve a purpose with proven, measurable and meaningful usage in the real world in terms of surge suppression. One thing it does not do is improve SQ nor does it claim to increase data transfer rates or fix packet errors (unlike the WW ETH Cables claims). Anyway, my experiment with using the Black Box ended over two years ago so your spelunking for dirt to throw has proven to be a waste of time.

 

[jkeny]

Is this thread about Wireworld or " How COULD upgraded Ethernet cables make a positive difference? What's behind it?"?

Yes, Amir, it's a complete falsehood & is a strawman as far as the thread question is concerned - you post a falsehood from a company & then provide data which has some bearing in disproving it - that's some sort of fallacious argument.

 

[jkeny]

I fail to see how the use of an Opto Isolator applies here. At least the Opto Isolator does serve a purpose with proven, measurable and meaningful usage in the real world in terms of surge suppression. One thing it does not do is improve SQ nor does it claim to increase data transfer rates or fix packet errors (unlike the WW ETH Cables claims). Anyway, my experiment with using the Black Box ended over two years ago so your spelunking for dirt to throw has proven to be a waste of time.

Why did you decide to use an opto-isolator on an ethernet cable - isn't the function of an ethernet cable just to deliver data?
What possible role can an opto-isolator have in this function?

PS. You really don't see the connection between using an opto-isolator in an ethernet connection - what the product you used stated "Protect equipment from being harmed by damaging ground loops caused by a difference in the ground potential of two connected devices." & what is being suggested might well be the reason for differences between ethernet cables & their carrying noise between devices?

 

[SoundAndMotion]

By the way, I was not necessarily asking for measurements. You mentioned there would be audible effect so I was interested in any listening tests that had shown that. Is there any such data?

Private listening data, yes, absolutely!

You said there was investigation going on. I am asking who is doing the investigation and the status of it. Does this question have an answer?

I already stated where this research is at but it appears that you don't seem to understand?

Hi John, I’m sorry, but I don’t understand. Are you saying that it is too early for research, and it hasn’t started yet? Or are you saying that the research is at a “private listening” stage…. and… ...you have done some? Are you aware of others also doing so?

The question of whether analog noise COULD be carried by an ethernet cable and COULD therefore interfere with a subsequent analog device (a DAC output amp, a preamp or power amp). It would seem that any designer/engineer would try very hard to block such a transmission path, but I know from experience (not audio) that grounding/shielding/noise problems can seem indomitable.

So, since it COULD happen, what do you know about further investigations?

Strawman? No, I am going by specific statements made by wire world about their Ethernet cable:

Hi Amir, I’m sorry but if only you and a couple of posts from cjf even mention data loss/corruption or packet loss/retransmission, how is it not a:
straw man
|?strô ?man|
noun
1. an intentionally misrepresented proposition that is set up because it is easier to defeat than an opponent's real argument.

Whose post(s) in this thread are you answering with your packet table? Everyone else is talking about (analog) noise… or did I miss something? If so, I apologize and would be grateful for the correction.

 

[amirm]

Hi Amir, I’m sorry but if only you and a couple of posts from cjf even mention data loss/corruption or packet loss/retransmission, how is it not a:
straw man
|?strô ?man|
noun
1. an intentionally misrepresented proposition that is set up because it is easier to defeat than an opponent's real argument.

Whose post(s) in this thread are you answering with your packet table? Everyone else is talking about (analog) noise… or did I miss something? If so, I apologize and would be grateful for the correction.

Hi there. Here is the thread of discussion:

I'm not a believer in fancy Ethernet Cables but I would be interested in seeing how the Ethernet traffic flowing thru these Cables look compared to a bog standard Eth cable. Are there any TX/RX errors, Discards...etc, are these Cables made to Spec in terms of their Impedance, does the fact that some of them use a totally different approach to how the wires are arranged within the jacket change their Spec compared to a "certified" bog standard ETH cable?

Just a few passing thoughts as a sip my wine and enjoy a night of Disco music in my favorite chair

The question was asked and I answered it with an explanation:

In such a short cable, you have to really, really go out of your way to corrupt any data. Rated at 100 meters/330 feet, a 3-6 foot section has way high of a signal to noise ratio to suffer from any ills.

I am going to go back to sipping my Earl Gray tea just the same.

If everyone was in agreement then that should have been the end of the discussion. But Yash posts this, quoting my post above:

Are you sure this is what Amir thinks? Because he writes the same erroneous assertion that it is about the data here:

He is saying I made an erroneous assertion. I took that as meaning that there are indeed data errors so I proceeded to provide the statistics.

Reading it back now, it is possible he is just making a snide remark that I am an idiot thinking data errors are a problem. But why would he say that when I am saying it is not???

Either way it is good to have real data that demonstrates lack of packet errors. A lot of people assert that but don't back it up.

This is not at all used by me as a strawman. I know the argument of noise and have asked for data on who has measured it way back in the thread. Nothing about answering Carl's question has anything to do with dismissing that argument.

 

[jkeny]

Hi John, I’m sorry, but I don’t understand. Are you saying that it is too early for research, and it hasn’t started yet? Or are you saying that the research is at a “private listening” stage…. and… ...you have done some? Are you aware of others also doing so?

Hi S&M - I thought I had explained it already but maybe not? I will talk about USB cables but consider it can be extrapolated to ethernet cables too - as I said before, both are packet based protocols & at the electrical level both use differential signalling but ethernet uses a signal transformer. So with those provisos, I consider the theory of what's happening on USB cabling to be useful for consideration in the context of ethernet cables. Remember this thread is titled " How COULD upgraded Ethernet cables make a positive difference? What's behind it?" so postulated mechanisms are relevant & on point.

I consider anecdotal evidence & personal observation to be a valid start to research of any topic. I can tell you that I & a local audio group have evaluated USB isolators & USB regenerators/reclockers & found both to improve the sound in many different configurations. I have experimented with USB isolation & regeneration & my results (verified in group listening) are again that isolation improves the sound but that a USB reclocker after isolation further improves the sound.

With these observations comes the simplistic postulates that some form of noise is being reduced/blocked by the isolation & that the waveform structure of the differential signals has still a possible role in audibility. Now delving further into these simplistic/summarised possibilities results in many questions that need to be addressed:
- is this common mode noise or is it differential noise that gets converted to CM noise at the differential receiver - or a combination of both?
- is signal waveform crucial to the behaviour of the differential receiver? For instance does a well formed differential (balanced) signal create a different noise profile on the ground of the receiver? Can this affect downstream circuitry to the receiver & result in audible issues?
- is it a ground noise issue or noise riding on the signal wires? I can tell you that when any & all ground wire issues are sorted that it is noise on the signal wires

And from these many more questions related to cables - can the balance of the differential signal be affected by the cable - the conformity of the twist, the configuration of the wires, the induced noise on the signal wires by the ground wire/shield, etc.

All of this is ripe for further investigation/experimentation which will take quite a lot of expertise & time to do. Most people are just interested in personal evaluation of the effectiveness of said approaches & make a decisions on that basis. It's a question of how one uses one's limited energy in this area.

I have done what I consider my due diligence in the area of USB transmission & have come to my conclusions but with open questions as to the exact mechanism of how all this ties together to produce audible differences. One of the factors in my understanding in all of this is, as you know, the workings of auditory perception & my conclusions that what we are seeing reported as audible improvements from much of these new initiatives in digital signal transmission, are more about the realism of the resulting soundscape rather than about individual distortion or frequency/amplitude issues. Hence I'm of the opinion that some fundamental but low level factor is being addressed in these audible improvements & that these elements are & will prove difficult to measure/isolate in the analogue signal.

The question of whether analog noise COULD be carried by an ethernet cable and COULD therefore interfere with a subsequent analog device (a DAC output amp, a preamp or power amp). It would seem that any designer/engineer would try very hard to block such a transmission path, but I know from experience (not audio) that grounding/shielding/noise problems can seem indomitable.

So, since it COULD happen, what do you know about further investigations?

As far as I'm concerned, we need a new measurement approach to answer the questions posed above & I do not have the expertise necessary to do this. It will take someone who is interested in this & who has the skill set to delve further & unfortunately this confluence of attributes only exists in a couple of designers, AFAIK .

Rob Watts of Chord, for instance & his findings/reports are interesting & relevant. His thoughts & investigations are to be found on Head-fi where he posts & prove to be an interesting read. He has come to similar conclusions about noise floor modulation & it's effect on the soundstage - see here & do a serach for him on Head-fi.

Similarly, John Westlake, audio designer of DACs has shown a measurement of USB signals (posted on PinkFishMeda forum) where the waveform was far off it's acceptable trace path at infrequent intervals (runt events) - this in a computer/laptop that was transmitting data without issue. After a USB regeneration device his plot showed all these runt events missing i.e. the waveform was well behaved.

PS. Oh forgot to mention Mallinson who was chief design engineer at ESS - there were youtube videos of his exposition of noise floor modulation & it's audibility at levels he couldn't hear it but others could in formal listening tests

Amir knows of this as I have previously posted these plots on his WBF thread about the Regen.