Network switches and cables

[davewantsmoore]

2 hours ago, El Tel said:

If you decide on the shielded, congratulations, you've used a cable intended only for data centre use and circumvented Ethernet mechanisms to prevent noise from propagation.

Indeed.

 

.... but to be balanced I think there's some important caveats here.

 

We cannot say that this noise is necessarily actually occuring, or is actually a "problem"......  so we cannot simply conclude that "whatever it is you are hearing is definitely due to the shield".    It depends on how the equipment is wired (it may be designed to cope just fine), and whether that noise gets anywhere near audio.

 

 

That being said, as a general statement, using a shielded cable probably isn't a good idea (for the reason you raise).

 

 

[NonPlayableCharacter]

3 minutes ago, davewantsmoore said:

We cannot say that this noise is necessarily actually occuring, or is actually a "problem"...... 

 

Hence in the opening part of my post I did not say it would create an issue, but that it can:

Quote

Use of shielded means all sorts of subtle distortions can start to impact upon the streamer/DAC

 

Your comment on beneficial distortion in audio being a myth is somewhat juxtaposed with how valves serve-up their output and are enjoyed by many, myself included. It is worthwhile differentiating this type of distortion from some whopping 50Hz harmonics caused by mains issues, just by way of example; but I think I get your point though.

 

[davewantsmoore]

1 hour ago, El Tel said:

Your comment on beneficial distortion in audio being a myth is somewhat juxtaposed with how valves serve-up their output

It's a myth.

Disclaimer:  I do own 7 valve amplifiers.... and my favourite of these (300B driven by 6C45P thru transformer) I used to think did sound this way (thick and warm), but... it's not.

 

 

The non-linear distorion isn't "benficial"..... and to confirm this you can mix in distortion of your own chosing (the same as an amplifier) to some audio, and see that below a reasonable number of %, it is quite inaudible.    

 

BUT... the distortion "profile" (ie. H2, H3, etc.) tells us a lot about the cause of the distortion in the amplifier ...... and those causes are very relevant to how an amplifier will sound under overload.... not because of the distortion it makes, but because of the recovery time and intermodulation products (low).

 

So quite the opposite of popular belief (tubes sound good cos distortion).... I think it's quite the opposite (tubes sound good because low distortion).

 

1 hour ago, El Tel said:

It is worthwhile differentiating this type of distortion

 

I honestly think that all examples of beneficial distoriton are mythology.

 

I do realise there are amplifier designers who think they can hear the "harmonic profile" of their non linear distortion .....  but IMVHO both theory AND practise (controlled testing) says they can't.    But that's not at all to say their is no benefit to the amplifiers they design.... the distortion is just a proxy (symptom).

 

The same goes for non lineardistortion products in DACs.

[NonPlayableCharacter]

58 minutes ago, davewantsmoore said:

It's a myth.

Disclaimer:  I do own 7 valve amplifiers.... and my favourite of these (300B driven by 6C45P thru transformer) I used to think did sound this way (thick and warm), but... it's not.

 

 

The non-linear distorion isn't "benficial"..... and to confirm this you can mix in distortion of your own chosing (the same as an amplifier) to some audio, and see that below a reasonable number of %, it is quite inaudible.    

 

BUT... the distortion "profile" (ie. H2, H3, etc.) tells us a lot about the cause of the distortion in the amplifier ...... and those causes are very relevant to how an amplifier will sound under overload.... not because of the distortion it makes, but because of the recovery time and intermodulation products (low).

 

So quite the opposite of popular belief (tubes sound good cos distortion).... I think it's quite the opposite (tubes sound good because low distortion).

 

 

I honestly think that all examples of beneficial distoriton are mythology.

 

I do realise there are amplifier designers who think they can hear the "harmonic profile" of their non linear distortion .....  but IMVHO both theory AND practise (controlled testing) says they can't.    But that's not at all to say their is no benefit to the amplifiers they design.... the distortion is just a proxy (symptom).

 

The same goes for non lineardistortion products in DACs.

 

One of the best thought-out and explained posts I have seen in a long time. Thanks for taking the time to do so.

 

There is a lot for me to unpack here and I'll try to bite off chunks and get my head around it at my own pace rather than blurt out any response without giving it the time it needs to sink in.

 

Am I right in my preliminary interpretation of your post that your assessment is that my enjoyment of valves over solid state is not just subjective, but is a psychoacoustic effect too? It's a strange one to come at if I can tell the difference immediately (blind) between a pair of JJ modern 12AX7s and a pair of NOS 1948 GE KenRad 12AX7s. Or am I missing the point?

 

We're getting way off topic here, but there is a thin thread of relevance with regards to the shielding conundrum.

[davewantsmoore]

1 hour ago, El Tel said:

There is a lot for me to unpack here and I'll try to bite off chunks and get my head around it at my own pace rather than blurt out any response without giving it the time it needs to sink in.

Have a look for papers and tests on the audibility of non-linear distortion.   It is just not very audible.

 

1 hour ago, El Tel said:

Am I right in my preliminary interpretation of your post that your assessment is that my enjoyment of valves over solid state is not just subjective, but is a psychoacoustic effect too?

No.

... at least not necessarily.

 

That might be why you experience what you do.... but I don't have enough information to say.

 

Your tube amp might actually be performing better overall than the amps you compared it to.... even though on a unclipped THD/IMD magnitude it is higher.

 

Also.... your tube amp may be producing a different frequency response at the speaker, than your other amplifiers.   Contrary to non-linear distortion (adding new frequency components to the signal that weren't in the original) ... linear distortion (changing the level of the existing signal components) is often much more audible, both becasue it occurs over a wider bandwidth... and because it is just much larger (remember a dB is very large).

 

1 hour ago, El Tel said:

It's a strange one to come at if I can tell the difference immediately (blind) between a pair of JJ modern 12AX7s and a pair of NOS 1948 GE KenRad 12AX7s. Or am I missing the point?

 

If you can tell the difference between them easily... then there will be something relatively trivial meaurable between the speaker output for each.    But not non-linear distortion, as this is not very audible (understatement).

 

1 hour ago, El Tel said:

We're getting way off topic here, but there is a thin thread of relevance with regards to the shielding conundrum.

Barely 

[NonPlayableCharacter]

5 minutes ago, davewantsmoore said:

Have a look for papers and tests on the audibility of non-linear distortion.   It is just not very audible.

 

No.

... at least not necessarily.

 

That might be why you experience what you do.... but I don't have enough information to say.

 

Your tube amp might actually be performing better overall than the amps you compared it to.... even though on a unclipped THD/IMD magnitude it is higher.

 

Also.... your tube amp may be producing a different frequency response at the speaker, than your other amplifiers.   Contrary to non-linear distortion (adding new frequency components to the signal that weren't in the original) ... linear distortion (changing the level of the existing signal components) is often much more audible, both becasue it occurs over a wider bandwidth... and because it is just much larger (remember a dB is very large).

 

 

If you can tell the difference between them easily... then there will be something relatively trivial meaurable between the speaker output for each.    But not non-linear distortion, as this is not very audible (understatement).

 

Barely 

 

Another great post. Thank you. I'll not derail the thread any further on this line of thought, but your structured post makes it easier to digest and it also points to the one component that pretty much everyone can actually agree on; that of system synergy being important.

[dbastin]

3 hours ago, davewantsmoore said:

I honestly think that all examples of beneficial distortion are mythology.

Umm, maybe except there is a long history of these ...

 

"It's mellow. It's raucous. It's tender. It's raw. It's the Maestro Fuzz-Tone. You have to hear this completely different sound effect for the guitar to believe it!"

 

[Monkeyboi]

On 28/08/2021 at 5:28 PM, Stereophilus said:

Pretty sure the fake Nordost Odin Cat 8 is the same as this AmpCom cable.

 

 

 

 Note how the person demonstrating the termination only performs a wire mapping test?  The REAL test of if you've done it correctly and the performance of the cable comes from the additional tests that the Fluke tester can perform such as resistance balance, transmission bandwidth performance, near end and far end crosstalk measurements and propagation delay (skew) testing.  Wire mapping alone tells you nothing about the quality of the cable, the connectors or the workmanship.

 

[BugPowderDust]

I do note with amusement that it is exceedingly rare that audiophile Ethernet cables post certification specification test results.
 

I imagine many would fail the fully array of tests to demonstrate their compliance with their nominated / advertised spec. 

[Stereophilus]

4 hours ago, BugPowderDust said:

I do note with amusement that it is exceedingly rare that audiophile Ethernet cables post certification specification test results.
 

I imagine many would fail the fully array of tests to demonstrate their compliance with their nominated / advertised spec. 

Indeed they may.  But the point is mostly irrelevant.  In the specific application of audio we only care that the data arrives at the reciever and if the sound quality changes.  If the former condition is met (it is a yes/no answer) then only the latter is in question.  

 

For the purposes of our discussion here, I guess the category designation really only serves to tell us what type of shielding (if any) is present.  But most IT specialists won't be using them, and most audiophiles won't care about the category designation (they will judge the cable with their ears)... so does it really matter?

[Monkeyboi]

4 hours ago, BugPowderDust said:

I do note with amusement that it is exceedingly rare that audiophile Ethernet cables post certification specification test results.
 

I imagine many would fail the fully array of tests to demonstrate their compliance with their nominated / advertised spec. 

 

Absolutely.   I also find it somewhat amusing that there's a lot of hype around Cat8 cables for audio which will invariably be connected to Gigabit router ports and switches which are more than likely to be equipped with Cat6 keystone sockets.   To realise Cat8 performance, ALL the components (sockets, cable, patch cords etc.) must be of Cat8 standard and both installed and fitted off correctly, then tested to certify it actually achieves that standard. 

 

It used to be a standard competency test for students undertaking their structured cabling and testing assessment to sneak a Cat5 socket into a network or have a concealed cable with a kink in it (minimum radius bending violation) to see if the student was performing the tests correctly and able to diagnose and locate the problem.  How many Cat8 installation links done by the ill-informed are longer than 30 metres?   I wonder how many unqualified DIYers doing their own installations actually know about these issues?  But hey 9/10 audiophiles will claim that measurements are meaningless, right?

 

Structured cabling when done correctly is actually very electrically robust.  IME and IMHO, a correctly installed Cat6 installation is overkill even for the highest bit rate audio.  Channel lengths (Link plus patch cables) can be as much as 100m @ 20 deg. C without the need for repeaters which should be more than long enough for domestic installations.

 

The most common installation faults I see in audits are -

• incorrect fit off of sockets (not maintaining the natural twist of the pair up to the termination)
• split pairs
• poor quality terminations resulting in resistance imbalances in the pairs
• incorrect shielding termination (if used)
• cable bending (too tight a bend in the cable)
• stretched cable (usually because of excessive hauling tension)
• using a lower category component in the link (e.g. Cat 5 socket with Cat6 cable in a Cat6 installation or a Cat5E patch cable used in a Cat6 installation).
• insufficient spacing between the cable and other services (mains wiring, switch-mode PSUs in ceilings, transformers etc)

All of the above impact on transmission quality.  You can use the highest quality components but if the workmanship is poor, the cable is incorrectly installed or there are a mixture of parts (e.g. mixing Cat5 with Cat6 components) the system will perform at best to the lowest common denominator.

 

 

[bob_m_54]

47 minutes ago, Stereophilus said:

so does it really matter?

 

Yes it does, if a supplier is selling the cable as a CatX device, and it doesn't conform. It makes a mockery of even labeling the cable as Cat anything.

 

Oh yeah, and it is illegal as far as misrepresentation of the item being sold..

[NonPlayableCharacter]

37 minutes ago, Monkeyboi said:

 

Absolutely.   I also find it somewhat amusing that there's a lot of hype around Cat8 cables for audio which will invariably be connected to Gigabit router ports and switches which are more than likely to be equipped with Cat6 keystone sockets.   To realise Cat8 performance, ALL the components (sockets, cable, patch cords etc.) must be of Cat8 standard and both installed and fitted off correctly, then tested to certify it actually achieves that standard. 

 

It used to be a standard competency test for students undertaking their structured cabling and testing assessment to sneak a Cat5 socket into a network or have a concealed cable with a kink in it (minimum radius bending violation) to see if the student was performing the tests correctly and able to diagnose and locate the problem.  How many Cat8 installation links done by the ill-informed are longer than 30 metres?   I wonder how many unqualified DIYers doing their own installations actually know about these issues?  But hey 9/10 audiophiles will claim that measurements are meaningless, right?

 

Structured cabling when done correctly is actually very electrically robust.  IME and IMHO, a correctly installed Cat6 installation is overkill even for the highest bit rate audio.  Channel lengths (Link plus patch cables) can be as much as 100m @ 20 deg. C without the need for repeaters which should be more than long enough for domestic installations.

 

The most common installation faults I see in audits are -

• incorrect fit off of sockets (not maintaining the natural twist of the pair up to the termination)
• split pairs
• poor quality terminations resulting in resistance imbalances in the pairs
• incorrect shielding termination (if used)
• cable bending (too tight a bend in the cable)
• stretched cable (usually because of excessive hauling tension)
• using a lower category component in the link (e.g. Cat 5 socket with Cat6 cable in a Cat6 installation or a Cat5E patch cable used in a Cat6 installation).
• insufficient spacing between the cable and other services (mains wiring, switch-mode PSUs in ceilings, transformers etc)

All of the above impact on transmission quality.  You can use the highest quality components but if the workmanship is poor, the cable is incorrectly installed or there are a mixture of parts (e.g. mixing Cat5 with Cat6 components) the system will perform at best to the lowest common denominator.

 

 

 

No argument whatsoever here (retired data centre designer and one-time BICSI certified cable-monkey here), however, I feel it is worth pointing out that there are many people happy to be wrong and not know why either. There are so many frivolous, extravagant and preposterous claims made on cables and switches that no matter how right you might be, it is just pearls before swine. 

 

As one learned individual once said to me: "I mostly ignore, but throw the occasional rock when I can be bothered".

 

It is good to see someone else stamping out the fires of ignorance to prevent those on the periphery getting burned though; I warn that it can be a lonely vigil sometimes.

[Stereophilus]

3 minutes ago, bob_m_54 said:

 

Yes it does, if a supplier is selling the cable as a CatX device, and it doesn't conform. It makes a mockery of even labeling the cable as Cat anything.

 

Oh yeah, and it is illegal as far as misrepresentation of the item being sold..

Is that not just a little petty?  Will anybody actually care?  We are talking home audio here...

 

And as far as misrepresentations go, surely the designated cat of the cable is the least of the audio ethernet cable makers sins against humanity... assuming of course someone can prove they don't meet the spec?

[davewantsmoore]

20 hours ago, dbastin said:

Umm, maybe except there is a long history of these ...

Of course (I am guitarist).

 

That distortion is much larger in magnitude..... but also much different in time ie. the amplifier (purposefully) doesn't recover as fast.

 

Also (even if they were the same magnitude) the masking effects in the audibility of non-linear distortion apply much much more to typical music .... than they do to "just" as guitar.

[bob_m_54]

15 minutes ago, Stereophilus said:

Is that not just a little petty?  Will anybody actually care?  We are talking home audio here...

 

And as far as misrepresentations go, surely the designated cat of the cable is the least of the audio ethernet cable makers sins against humanity... assuming of course someone can prove they don't meet the spec?

The problem is, someone may actually connect the cable into a system that is not their Home Network. And if a manufacturer labels the cable as a CatX, then it must meet the CatX compliance requirements. Hell someone may wan't to use the cable in a system where data transfer capability is actually critical.

 

Now if the Manufacturer wanted to cover their butt (or the importer) then they can add another couple of labels to the cable. Scroll to the bottom of this page:

https://www.accesscomms.com.au/ref_comply/

 

Edit: Or just market the cable as an Audiophile Network Cable, and not claim that it complies with any standard.

 

Edited March 19, 2022 by bob_m_54

[davewantsmoore]

22 hours ago, El Tel said:

Another great post

I've put a lot more thought into others, that didn't get such a nice reaction, so cheers.

 

22 hours ago, El Tel said:

system synergy being important.

Oh dear lord, don't get me started again  

 

Suffice to say that I think that with this one, some actual real, performancy things, can be hidden behind it ..... that a lot of the time it doesn't serve the general forum consumer well.

[NonPlayableCharacter]

12 minutes ago, Stereophilus said:

Is that not just a little petty?  Will anybody actually care?  We are talking home audio here...

 

And as far as misrepresentations go, surely the designated cat of the cable is the least of the audio ethernet cable makers sins against humanity... assuming of course someone can prove they don't meet the spec?

 

If you're claiming a CAT rating then it has to make the specs. Otherwise it is outside of category and non-compliant and should come with a disclaimer explaining its intended use and/or prohibitions. It matters.

 

Spurious claims abound. One maker, for example, claims one of their Ethernet cables is CAT9. There is no TIA standard for CAT9.

[davewantsmoore]

1 minute ago, bob_m_54 said:

if a manufacturer labels the cable as a CatX, then it must meet the CatX compliance requirements.

 

Actually much harder than it might be assumed to make a cable that passes in higher categories.

 

Kinda not surprising if you consider that like HDMI we're running very high frequencies over twisted pairs, that are sometimes even terminated with the sorts of push-in / crimp "field assembly" style plugs.

 

 

 

[Assisi]

On 17/03/2022 at 6:32 PM, davewantsmoore said:

The way I read it El Tel has been saying that it could never affect the data.... which is correct.

 

As for whether anything ELSE could affect the resulting audio ..... whether that is going to happen depends.    Think about it.... this is why you get many people reporting different things.... and many people reporting nothing.    Because if you think about how those mechanisms could affect a system, you would never expect it to be the same on every piece of equipment, or to be a problem on every/any combination of equipment.

• Noise in the shield, that is then inadvertenty mixed in with a (digital or analogue) audio signal.
• Packet jitter that inadvertently leaks through the Ethernet PHY into digital audio signal timing.
• Noise getting onto the mains, making its way into another audio device.

 

This is not something that manufacturers typically don't give you a lot of insight into.... and that most other discussions frame into an "always vs never", "believers vs naysayers" type of thing.

 

It certainly happens though.....  it's obviously why the people trying to build and improve top flight gear do extremely agressive things to avoid the above issues from having their way.

 

I agree that the data will be correct.  I am certain that we all listen to exactly the same streamed “music”.  The difference is that we all listen to different levels of noise.  I consider that noise can spoil the listening experience. 

You have mentioned and acknowledged the existence of “noise” in this post and another recently.  In the context of “noise in the shield”, what noise(s) are you referring too?  What causes it or them?  Where does it or they come from?  As well as using unshielded cables, what other strategies can be implemented to mitigate the existence of the noise getting into and going through a shielded cable? 

If CAT6 or 7 cabling is used with metal plugs and the shield is not connected at one end or both will that solve any potential problem for the  shield to transmit noise?

John

[bob_m_54]

16 minutes ago, davewantsmoore said:

 

Actually much harder than it might be assumed to make a cable that passes in higher categories.

 

Kinda not surprising if you consider that like HDMI we're running very high frequencies over twisted pairs, that are sometimes even terminated with the sorts of push-in / crimp "field assembly" style plugs.

 

 

 

Actually no, it just has to meet and be certified to comply with the requirements.. That's what Certification of Compliance is all about.

[Niktech]

49 minutes ago, Stereophilus said:

 

 

And as far as misrepresentations go, surely the designated cat of the cable is the least of the audio ethernet cable makers sins against humanity... assuming of course someone can prove they don't meet the spec?

If an allegation is made one would need to prove the nature and extent of the misrepresentation. 
 

if the cable does in fact meet spec and no misrepresentation is proved, a claim for defamation can be made against the person making the allegation because the maker/manufacturers  reputation has been harmed, including damages for the loss of sales

Edited March 19, 2022 by Niktech

[Monkeyboi]

48 minutes ago, Stereophilus said:

Is that not just a little petty?  Will anybody actually care?  We are talking home audio here...

 

And as far as misrepresentations go, surely the designated cat of the cable is the least of the audio ethernet cable makers sins against humanity... assuming of course someone can prove they don't meet the spec?

 

Well actually measuring the performance of Ethernet cables isn't rocket science.  With a bit of knowledge, the correct measuring equipment and techniques it only takes a few minutes to readily confirm if the cable or the installation meets the required standard.  If it fails just one of the tests then it doesn't comply with the standard.  It's cut and dried.  Simple as....

 

That's why we have compliance testing standards.  A manufacturer can label anything they want on a cable or any other piece of equipment for that matter and make as many phantasmagorical claims as they like, but if it doesn't measure to the acclaimed specification it's misrepresented.

[davewantsmoore]

26 minutes ago, bob_m_54 said:

Actually no, it just has to meet and be certified to comply with the requirements.. That's what Certification of Compliance is all about.

I said it was more difficult than people might assume to make a cable which passes higher cat tests.

 

... yes.   "it just has to comply with the requirements".   What's your point?  What did I say that was wrong?

[davewantsmoore]

22 minutes ago, Niktech said:

If an allegation is made one would need to prove the nature and extent of the misrepresentation. 

Some years ago there were a few sellers claim to have Cat8 cables... when such a category didn't even exist.

 

Now apparently the same is happening with 9.