Archimago measures Ethernet cables

[aussievintage]

43 minutes ago, Stereophilus said:

Extrapolating beyond the external validity of the DBT they are quoting to suggest that certain things are inaudible, certain things are a waste of money, or both.

 

Not just DBTs.  People extrapolate all sorts of notions trying to validate their own ideas.   

[sir sanders zingmore]

4 hours ago, davewantsmoore said:

I still don't understand.

 

The "WTF MOIT" highlighted areas are identical on each of the 7 tests.    What are you saying?   They're not actually identical?.... or that a different stimulus needs to be used? (which, if so, you just appeared to contradict)

 

Remember - we are only talking about jitter caused/causable by the change of cable.

Just returning to the topic (Archimagos finding of no difference between the cables he tested.)and trying to ensure that Dave’s excellent point doesn’t get lost in the ‘weeds’ of the discussion. 
 

 

[Satanica]

3 hours ago, Stereophilus said:

Control introduces bias.  Even when done “correctly” there is bias.

 

Let’s use Toole as our example.  In one of his DBTs comparing speakers he controls for listener intervariability by training his test subjects to listen for (I think it was) 6 different facets of the reproduced music.  Only the 6 that Toole felt were most important.  That is bias.

 

Unless I'm mistaken in the same test(s) there were untrained users as well.
And the trained non-professional reviewers did much better at scoring speakers blind versus sighted compared to the non-trained professional reviewers.

So just imagine if there were more than just 6 facets, perhaps they would have smashed the non-trained professional reviewers even more?

Edited March 20, 2021 by Satanica

[Guest Eggcup the Dafter]

45 minutes ago, Stereophilus said:

You are referring to validity.  Yes, the internal validity of a DBT maybe intact but the point of doing a DBT is to gain external validity.  That is, you want to draw broader conclusions about what you are studying, otherwise why test at all.

 

So, by all means use a DBT to prove that you can train people to pick certain differences in speakers.  But the bias inherent in doing so severely limits the application of your DBT conclusions in a more general context.  And unfortunately we DO see people in many forums doing just that.... Extrapolating beyond the external validity of the DBT they are quoting to suggest that certain things are inaudible, certain things are a waste of money, or both.

It's clear from this statement that you don't understand what a DBT is or what it does. If the other side don't understand it either, that does not make the actual process invalid.

The point of doing a DBT is tp prove a single hypothesis, nothing more.

If you want to support a wider theory with DBTs you have to conduct a series of them to support that point.
 

It is exactly the same as conducting a sighted single subject test. The result applies to the specific item or system under review  - and the single subject. It is not universally or more widely applicable, and the subject is a key part of the test.
 

However, the subjectivist cult doesn't work on that basis. In essence, if somebody reports a difference from changing an Ethernet cable in one specific instance... it's now universally applicable. If someone else can't hear it they are cloth eared, their system is insufficiently resolving, etc.

It goes the other way as well. If you do hear a difference, the words "confirmation bias" will be thrown at you in no time, see above. I suspect it's more complicated than that.

 

The difference may or may not be there until it is proven, until then it is just two sides slinging mud.

So, back to the original subject. How do you prove that two Ethernet cables cause different soundwaves in the room? (You've proved already that a listener can "hear a difference" in a particular context, just by doing it yourself - but that's different).

 

Let's face it, the musical signal is extracted from the packet, buffered, probably decompressed from FLAC or a similar format, passed through another cable to another buffer, put through at least one and possibly several digital filters, before the actual conversion process in whatever form - maybe then passed through an analogue filter, several gain stages and cables, maybe tone and loudness controls along the way, more cables and finally some speakers turn it into sound. That's a hell of a lot of subsequent processing including multiple changed digital signals, while the Ethernet stage is proved "correct" in that it passed the music on for all that subsequent processing (if it breaks, no music!).Then at the end of it all, what - a fallible human reports a "change" so it must be there and it must be caused by the Ethernet cable, nothing else. Put that way, I'm sure you can see it's a fairly way out proposition, and it's one to be proved, not disproved.

What if the difference does take place in our brains? Well, I'll have some of that, please, as long as it doesn't cost as much as an Audioquest Diamond ethernet cable, turn me into an arch conservative, or cause cancer. It's just as valid as a new pair of speakers.
 

 

[davewantsmoore]

3 hours ago, Assisi said:

Has anyone poster in this thread done it correctly, whatever that may be?

 

Unlikely.... it's very difficult and tedious, and error prone.    It's only of any use at assessing how we hear and perceive..... not being used to assess specific audio components.

 

3 hours ago, Eggcup the Dafter said:

The thing about the DBT is that it’s the best we’ve got for proving a difference that we can hear. It’s not infallible,  but if you prove a difference in a properly run DBT then there is a difference.

 

Perhaps....

 

However .... Given the state of measurement (humans can measure mind-bending orders of magnitude away from 'audio') .... I find it most bemusing that we do not rely on actually measuring that the difference exists (eg. electrical, or pressure waves).

 

I find the notion of a listening test for a thing which I have not already quantified the existence of, completely insane.... and instead I'd prefer to take the advice of Mr Metal Fan.

[muon*]

I see plenty of biases in a lot of these posts about DBT's

[davewantsmoore]

3 hours ago, acg said:

The real question though, is who can pass a J-Test after a couple of drams?  Not me occifer...

 

I was always told....  beer then grass you're on your arse..... grass then beer you're in the clear.

 

I'll pass (the test) either way  .... although I'd prefer it to be sipn drams, as beer doesn't do me any favours these days.

[Stereophilus]

1 hour ago, aussievintage said:

 

Not just DBTs.  People extrapolate all sorts of notions trying to validate their own ideas.   

Indeed they do.

 

And to segue things back on topic, your statement is true of the Ethernet cable discussion:

 

Archimago’s testing is interesting and useful, but cannot, IMO, be more broadly used to justify a position that different Ethernet cables make no difference to sound quality when used in an audio application.  Using that same logic, I do not believe every fellow audiophile who tells us Ethernet cable A sounds better than Ethernet cable B, because that is what they heard.  And, I do not believe marketing claims that Ethernet cable X must sound better because of its unique conductor properties (or whatever).  
 

What I do, which works for me, is I evaluate all of the information available to me from the above sources.  I consider it and I use it to make a decision on whether or not to try something in my system.  My final critical appraisal test is my own listening, because I’m buying for only one person: Me.  I don’t own test bench equipment, so I use my own ears, in full knowledge that I have bias, my hearing isn’t perfect and that I am a sample size of 1 for statistical purposes.

 

I actually tried 6 different Ethernet cables in my system last year. None made any difference whatsoever.  Some were fancy and expensive, some were cat 8, one was cheap Bunnings cat 5e.  I did recently come across another different Ethernet cable recently.  This one does make a difference.  Quite subtle, not glaringly obvious, but consistently reproducible.  It has given me cause to become interested in why and how.

 

46 minutes ago, Eggcup the Dafter said:

It's clear from this statement that you don't understand what a DBT is or what it does. If the other side don't understand it either, that does not make the actual process invalid.

No, I do.  What I said was in rebuttal to a very specific point.

 

Quote

The point of doing a DBT is tp prove a single hypothesis, nothing more.

If you want to support a wider theory with DBTs you have to conduct a series of them to support that point.
 

It is exactly the same as conducting a sighted single subject test. The result applies to the specific item or system under review  - and the single subject. It is not universally or more widely applicable, and the subject is a key part of the test.

Agreed.

 

Quote

However, the subjectivist cult doesn't work on that basis. In essence, if somebody reports a difference from changing an Ethernet cable in one specific instance... it's now universally applicable. If someone else can't hear it they are cloth eared, their system is insufficiently resolving, etc.

It goes the other way as well. If you do hear a difference, the words "confirmation bias" will be thrown at you in no time, see above. I suspect it's more complicated than that.

I think we both share the same distaste for evidence extrapolated beyond its context. 
 

Edited March 20, 2021 by Stereophilus

[Assisi]

The thread started with a link to a review that is of itself from another time on the matter of the impact Ethernet cables on digital streaming.  To me, the switches and the cables that the review is based upon were for just ordinary stock standard network intended purposes.

 

I am aware that a couple of the posters in the thread and myself have good audio intended network setups.  How many other posters have tested cables intended for audio or listened to, or experienced a system with a good quality audio network setup and perceived the benefits?  I wonder?  Just like other aspects of today’s audio digital strategies, the audio network world has moved on considerably since the review was written.

 

 

The thread has morphed into an opportunity for various posters that in the main too often express numerous sceptic comments on many aspects of things network and related.  It has wandered all over the shop sprinkled with a liberal use of I thinks, maybes, conjectures, assumptions etc.  Then there are the unnecessary put downs and insults from some posters.  Much of what has been said is not particularly useful to the actual topic.  The review is well out of date.

 

 

I have had considerable experience over the past year or so with my own audio network setup and others.  To me there are conclusive listening benefits from the use of quality Ethernet cables and other aspects of networking including switches for audio.  At a technical level I expect that everything with the transmission and receiving of the network information worked perfectly years ago and it is no better or different today.  For example, from my experience, for some reason that I cannot explain, two different audio switches or more are better than just one.  You get the combination of the signature outcome of each.  No DBT needed to discern the impacts of the combination outcomes.  It definitely happens.  Why? 

John

 

 

[acg]

I don't use a network John, unless I am streaming from Roon/Tidal, and then I just expect it to sound sub-par and have never really tried to get it up to speed.  Streaming is for background music here. 

 

That is not entirely true, there is a network  Tablet >> wifi >> router/switch >> wire >> switch >> wire >> MusicServer >> wire >> AudioPC.  Between MusicServer where my files are stored and the AudioPC which send the data to the dac is a direct ethernet cable connection with zero switches or other componentry.  The AudioPC grabs the file for playback from the MusicServer and then totally shuts own the GUI.  AudioPC is not internet connected and its only friend is the MusicServer.  Anyway, I bought a fancy shielded ethernet cable to go between the two because I am a little worried about ethernet galvanic isolation only being about 60% effective (if what I read on the interweb is true) and the eight computers in my office/listening room (nine including a RPi...but is that really a computer?) and approximately 300kg of magnetics in my amplifiers/preamp/phono etc even though the RFI/EM meter suggests all is golden in most parts of the room.

 

That was long-winded.  Basically, I bought this ethernet cable with adjustable shielding and have not bothered to even listen to see if it makes a difference to anything, let alone adjust the shielding to see if I can get it doing a better job, assuming that there is even an issue to solve in the first place.  There might be, but that is for investigation another day.

[aussievintage]

21 minutes ago, Assisi said:

sprinkled with a liberal use of I thinks, maybes, conjectures,

 

Partly because, around here, if you state something as a fact, you get shouted down, or at least come across as arrogant.  You have to be careful when you want to be absolute about something.

[Assisi]

11 minutes ago, acg said:

I don't use a network John, unless I am streaming from Roon/Tidal, and then I just expect it to sound sub-par and have never really tried to get it up to speed.  Streaming is for background music here. 

Given the attention to serious detail that you have applied to other aspects of your system maybe a bit of Network development and tweaking may be in order for your listening pleasure.  A good switch and cables plus LPS if needed could be of benefit for you with ROON/tidal so that it is not subpar.  With switches there are so many now whilst cables can be a bit of a conundrum.  You can only try.

John

[Guest Eggcup the Dafter]

36 minutes ago, davewantsmoore said:

 

However .... Given the state of measurement (humans can measure mind-bending orders of magnitude away from 'audio') .... I find it most bemusing that we do not rely on actually measuring that the difference exists (eg. electrical, or pressure waves).

 

I find the notion of a listening test for a thing which I have not already quantified the existence of, completely insane.... and instead I'd prefer to take the advice of Mr Metal Fan.

That's a fair point. I could (as usual) have been a bit more exact in what I wrote.

[Satanica]

What I don't undertstand is why there are ever posts with reports of sighted subjective listening impressions in a debate topic unless the person behind the posts is prepared to accept that they might be wrong.

[Guest Eggcup the Dafter]

43 minutes ago, Assisi said:

The thread started with a link to a review that is of itself from another time on the matter of the impact Ethernet cables on digital streaming.  To me, the switches and the cables that the review is based upon were for just ordinary stock standard network intended purposes.

 

 

I am aware that a couple of the posters in the thread and myself have good audio intended network setups.  How many other posters have tested cables intended for audio or listened to, or experienced a system with a good quality audio network setup and perceived the benefits?  I wonder?  Just like other aspects of today’s audio digital strategies, the audio network world has moved on considerably since the review was written.

 

 

 

 

The thread has morphed into an opportunity for various posters that in the main too often express numerous sceptic comments on many aspects of things network and related.  It has wandered all over the shop sprinkled with a liberal use of I thinks, maybes, conjectures, assumptions etc.  Then there are the unnecessary put downs and insults from some posters.  Much of what has been said is not particularly useful to the actual topic.  The review is well out of date.

 

 

 

 

I have had considerable experience over the past year or so with my own audio network setup and others.  To me there are conclusive listening benefits from the use of quality Ethernet cables and other aspects of networking including switches for audio.  At a technical level I expect that everything with the transmission and receiving of the network information worked perfectly years ago and it is no better or different today.  For example, from my experience, for some reason that I cannot explain, two different audio switches or more are better than just one.  You get the combination of the signature outcome of each.  No DBT needed to discern the impacts of the combination outcomes.  It definitely happens.  Why? 

 

John

 

 

 

So, if you're slightly concerned with why, my question still stands. How do you prove that the sound waves in the room have changed? Not what you or anyone else have "perceived" with sighted tests etc.

 

What you have proved is that you perceive a difference. That is what definitely happens.

 

What you have not proved is that the sound waves have been changed by the audio switches, cables or anything else. It's not "definite" at all.

So answer your own question, will you, before asserting that the change is "definite", please.

 

Why?

 

 

 

 

[Martykt]

3 hours ago, aussievintage said:

 

Actually, thinking about it, it depends what he set out to prove.  If he was trying to prove people could be trained to hear those facets, then maybe it's a valid test    Seriously, you have to design the test with the goals clearly in mind.

This in itself can be a bias as well.

When designing tests what a person designing the test wishes to prove or disprove can influence the results due to how the test is structured.

[sir sanders zingmore]

There's a lot of talk about the biases that DBTs can introduce and the fallibility of that method of testing.

We can discuss that for sure, but there cannot be a double standard that claims bias in DBTs and yet simultaneously claims or at least implies, that sighted listening is somehow not prone to bias

[acg]

Not sure anybody has asserted that @sir sanders zingmore.  As far as I am concerned we are all conscious and are therefore biased...just the nature of, well, nature.  

[rantan]

1 minute ago, acg said:

Not sure anybody has asserted that @sir sanders zingmore.  As far as I am concerned we are all conscious and are therefore biased...just the nature of, well, nature.  

 

100% agree.

[Satanica]

3 minutes ago, acg said:

Not sure anybody has asserted that @sir sanders zingmore. 

 

I've never been more sure about anything in my life.

Edited March 20, 2021 by Satanica

[sir sanders zingmore]

8 minutes ago, acg said:

Not sure anybody has asserted that @sir sanders zingmore.  As far as I am concerned we are all conscious and are therefore biased...just the nature of, well, nature.  

 

I would assert that when someone asserts that you should just "trust you ears"*, they are indeed asserting that ears can be trusted and are therefore not biased 

 

 

*paradoxically, "trust your ears" in this context means perform a sighted listening test rather than a test that involves only ears

[acg]

24 minutes ago, sir sanders zingmore said:

I would assert that when someone asserts that you should just "trust you ears"*, they are indeed asserting that ears can be trusted and are therefore not biased 

 

To quote Dr Maturin from the wonderful books by Patrick O'Brian, each of us must "choose between the lesser of two weevils".  Not everybody is comfortable with the same insipid insect...and that is fine.

[muon*]

1 hour ago, sir sanders zingmore said:

I would assert that when someone asserts that you should just "trust you ears"*, they are indeed asserting that ears can be trusted and are therefore not biased 

That is an assumption of course

 

Could be that they view it that way because at the end of the day we listen to our music with our ears.

 

Meh, different strokes, each to their own, who cares what someone else does when it is not what they choose to do, and if they do? why do they?

[Martykt]

1 hour ago, sir sanders zingmore said:

 

I would assert that when someone asserts that you should just "trust you ears"*, they are indeed asserting that ears can be trusted and are therefore not biased 

 

 

*paradoxically, "trust your ears" in this context means perform a sighted listening test rather than a test that involves only ears

Well actually yes, ears are basically operate like microphones.

It's how we interpret the signals that the ear puts out that's important.

It's important to realise that bias is psychological in nature and won't effect how the ear actually operates.

 

It's also important to understand the limitations of DBT's where the brain quite often needs context to properly process the information being fed it.

[davewantsmoore]

2 hours ago, Satanica said:

What I don't undertstand is why there are ever posts with reports of sighted subjective listening impressions in a debate topic unless 

 

2 hours ago, Eggcup the Dafter said:

Why?

 

Many/most people refuse to believe anything else than if it "sounds different", then it was actually different.......  which is understandable, when you ponder it.