Which ethernet cable?

Page 4 - Seeking answers? Join the What HiFi community: the world's leading independent guide to buying and owning hi-fi and home entertainment products.

ThisIsJimmy

Well-known member
Nov 11, 2020
292
122
2,070
Visit site
There have been a number of positive reviews of cables and switches from the HI-Fi press, including “What Hi-Fi”, owner of this forum over the years.

However, where they state a positive recommendation for an expensive cable or switch, how did they come to this conclusion.

The problem I have here is not that they recommend it because it makes an improvement, but what are they benchmarking against to begin with? What is the environment like in the house or testing environment? Where is the router/switch going to be located they are patching into? What's equipment/devices are around the switch or cable? Where does the cable need to travel between and what is the distance? What is the end device they are testing this with? What is the testing applications suite? Browsing through a bunch of reviews, I cannot find an example of where this is all provided.

A decent Ethernet cable manufacturer usually has a load of blurb on their cable which may look similar to the attached patch lead image which gives a very good indication about whether or not you are using the right cable in the correct environment.

If the benchmark a reviewer is testing against is an unbranded, unshielded CAT5 cable, that for example, travels near an amplifier, and they are testing using a UDP stream application, then the result of the performance comparison from what we have already discussed shoud be self explanatory and should not need a listening test to know which will be better.

Comparing the Excel Cable I have pictured against an Audiophile cable I just jumped onto future shop and picked out a random cable: https://www.futureshop.co.uk/audioquest-cat6-pearl-rj-e-ethernet-cable-custom

The initial statement falls in line with the CAT6 cabling standard - 10Gb to 55m
From the descriptions and specifications provided it suggests the outer sheaf is shielded, but doesn't not mention any shielding on the copper twisted pairs, so looks like a form of S/UTP cable (Shielded with unshielded twisted pair).
The silliest statement I found was that this is a "directional cable" and that "The correct direction is determined by listening to every batch of metal conductors"... From the videos I have already posted above about how computer communication works, i'll let you come to your own conclusions about that. As for the price, £17.50 vs £8.50 I found for the excel F/FTP (Foiled with Foiled twisted pairs) cable.

If you know what you are looking for here for starters, narrowing down any ethernet cable choice becomes much easier and would make any comparable review directly comparable, relative and informative.

If the switch makes that much of a difference, to me it really says something about an issue with the environment it is located in. What is around it? Sticking it on top of an amplifier of subwoofer is silly. Likewise near christmas tree lights or your microwave isn't a smart place either or near to cordless home telephone. If you router is performing badly or making a weird noise get your ISP to replace it. Finally, where is the UDP streaming source as this protocol is where the difference would be notable on any comparison.

If you know your are working an an EM heavy environment, then hardened industrial switch's are available, though like AV switches these too are very expensive as they are built for a purpose.
 

Attachments

  • Patch Lead Image.jpg
    Patch Lead Image.jpg
    681.9 KB · Views: 6
  • Like
Reactions: Symples

ThisIsJimmy

Well-known member
Nov 11, 2020
292
122
2,070
Visit site
No doubt the original XLRs were made with identical cable on both the left and right channels Jimmy.
So how could (working) cables have been the cause of any imbalance?

I think in this case it's not necessarilly the cable, but i'm wondering if the performance is because this isn't a pre-manufactured XLR-XLR cable, but made up cable order from ebay. The cabling these cables was made from, the Mogami Neglex 2534 which looks to be a popular cable. Is it worth getting these sent off to be reterminated?

Objectively in hindsight the only thing I haven't done is tried swapping the existing cables round.

Initially I wondered if:

1) It was my ears - I've always had issues with my right ear. Ruled this out when listening through my Impact Universal which is digitally channel balanced
2) It's also not headphone related because of the above
3) I wondered if it was the Volume POT on the Topping A90 - this is a rather common complaint about this Amplifier. Topping having just rolled out the A90 Discrete with full channel balancing, and this did have me wondering if it would be any better or not, or whether I should swap my XLR Interconnects out before making the plunge. I'm glad i did the latter. Even if it turns out now that this is the POT on the amplifier, it's not noticeable now with my new QED Reference XLR 40 2 XLR to 2 XLR Audio Cable Pair.
 

Attachments

  • PXL_20220630_190525266.jpg
    PXL_20220630_190525266.jpg
    100 KB · Views: 6
  • PXL_20220630_190624322.jpg
    PXL_20220630_190624322.jpg
    117.8 KB · Views: 6

Gray

Well-known member
I wondered if it was the Volume POT on the Topping A90
That seems more likely, though as you're suggesting, it wouldn't explain why the new XLR leads cured it.

I doubt there's anything wrong the e-bay seller's construction but, if you've got one, you could just use a multimeter's continuity check to confirm input to output pin 1 to 1, 2 to 2 and 3 to 3 on each of those original XLR leads.
 
Last edited:

RobGardner

Well-known member
Jul 22, 2008
45
14
18,545
Visit site
The stream over your Ethernet system will only be as good as the weakest link, so changing a single cable will probably make little difference. Even running a single upgraded cable directly between your router and streamer will likely not make much difference unless other factors are accounted for.
In a complex Ethernet system with switches, changing a few cables in key locations can improve audio quality (Just) But I have found that much more is needed to get significant gains in audio quality.. The good news is that if the whole system is optimised the benefits can be shared over more than one device and while not cheap as chips, with careful sourcing it needn’t cost a large ( in hifi terms) amount of your hard earned cash.
I have done the following with excellent results, but it was only once everything was in place that the audio and audio visual quality really took off to another level.

1 - Good quality Ethernet cables, I chose uGreen cat8 for the significant emi/ Rfi shielding and quality construction. I initially used these everywhere but now I use these mainly for the long runs. At key points I now have home made cat8 Ethernet cables using DH Labs reunion cable with Telegartner connectors. I would like more of these cables but can’t get hold of the cable off of the reel anymore and while they do bring improvements it’s not night and day.. I have tried various other high end ethernet cables and while there is sometimes a sound quality uplift, generally, it’s not worth the considerable amount of money for the tiny incremental gain in audio quality. With a sub optimal Ethernet loom I would expect even less uplift.

2 - Good quality switches, I have one Chord English electric audio switch and a home tweaked Netgear switch In my network. The tweaking only consists of applying Dynamat sound deadening foil to the inside of the switch. This tweak lifted the audio visual signal, presumably a combination of resonance control and some additional shielding afforded by the bitumin and aluminium foil. The difference was clearly audible. Initially I was very impressed with the Chord audio switch vs my tweaked switch but I now wonder if the main improvement was in the Teddy Pardoe power supply I used with the Chord. I would like to audition a super high end switch but I am getting such great results I don’t need to scratch that itch for the moment.
3 Excellent quality linear regulated power supplies for both the switches and router. I purchased a used dual output 12v regulated linear power supply on eBay and use this to power my Virgin Media router and the tweaked switch. As mentioned, I have a Teddy Pardoe 5v supply into my Chord switch.

With careful purchasing and using second hand parts, it has probably cost about £600 to do all of this, (including the new Chord switch) but the uplift in quality has been tremendous. If I now remove any single part of the network and replace with bog standard parts the whole system suffers. Interestingly, changing the home made Ethernet cables for the Ugreen cables has the least impact.
I recently borrowed Network Acoustics Muon Ethernet filter and cable. Used together from the Chord switch into my Devialet amp they really brought everything together, but the improvement was much much less than getting the network “right”. I might save up for a Muon, but it’s a lot of money to improve just one aspect of my system. I don’t want to be constantly moving it from my amp to av system as the need arises.

What I am saying in summary, is that I support that old noughties hifi cliché that you need to consider the whole wiring loom, in this case, of your Ethernet network to really sort things. If you are patient and can wait to buy kit at the right price it needn’t cost a fortune and the results are cumulative. In my experience just changing a single cable has negligible impact and you might be better off exploring other options.
 

djh1697

Well-known member
Nov 27, 2008
171
30
18,620
Visit site
Teddy Pardo does not have a dealer network, I emailed him to ask him about this, his reply was "dealers do not work for nothing" which is entirely correct. He offers a full refund if you are not happy, less p&p of course.
 

RobGardner

Well-known member
Jul 22, 2008
45
14
18,545
Visit site
Even with my failing hearing I can hear a clear difference between some Ethernet cables. Whether it’s the quality of the wire, the connectors or the shielding, they can and very often do sound different to one another. I have just replaced a 8m cat8 uGreen cable (which sounded very good) with something that sounds better but cost about 10 times the price. Generally I have been unimpressed with cheaper “audio” ethernet cables and swopped them out for a uGreen cat8 loom and used the extra money from selling my “audio” cables to buy something else.
Then, one day, my brother in law came around with a super high end Ethernet cable that had been universally very well reviewed. It “sounded” very good indeed but not as good enough to make me want to buy one, However, it did set me off on trying to achieve a cost effective upgrade that would go someway towards recreating the achievable audio improvements.
I like the sound of silver plated copper and pure silver conductors in audio connectors, so that’s what I looked for. I like value for money and I don’t mind getting my hands dirty, so I thought I might try making a cable. I noticed most high end Ethernet cables use Telegartner connectors. My initial thoughts were simply to try replacing the connectors on my uGreen cables with Telegartner connections and see if that improved things, but as the connectors cost more than the cables I thought I would go the whole hog. and change the cable as well. I then found DH Labs reunion cable was available off the reel., it was cat8 (best shielding spec) silver plated copper cable. I started with a making a single length, swopping out the final length of uGreen cable to my streamer and was initially not blown away by the very slight change. I left my system running over the next couple of days and presto! more detail and a generally smoother and less harsh sound quality. I have slowly changed the three lengths of cable between my NAS and streamer and each increment was a significant improvement. The final link was the long 8 m length. Not great value for money, but every little bit helps.
Making the cable was easy and I already had the tools needed. (stanley knife, wire stripper, mole grips, small adjustable spanner).
I would say, without trying to sound elitist, that unless you have a high end and super revealing system you will hear so little difference (if any) that you should stick with a good quality standard cat8 Ethernet cable and spend your money on other things Like a bottle of wine or putting the heating on.
 
I would say, without trying to sound elitist, that unless you have a high end and super revealing system you will hear so little difference (if any) that you should stick with a good quality standard cat8 Ethernet cable and spend your money on other things Like a bottle of wine or putting the heating on.
Thanks for the write up. For context, what streamer do you use, and have you a network switch either just after the hub/router or before the streamer? Or maybe both?

In my case, my signal goes:- FTTP > ISP‘s generic Huawei Router > 1m Cat6a > Netgear GS108 switch > 10m Cat 6a > Linn streamer, per signature. The cables are Duronic FTP Cat 6a generic ones.

I’ve read that some suggest a switch within a metre or so of the streamer gives the best isolation, but I’ve not tried that yet. Re the Ethernet cable, I favour reasonable flexibility and price over material content in the belief it either gets through or it doesn’t. Besides, Linn don’t suggest higher grade or foo cables.
 
Last edited:
  • Like
Reactions: WayneKerr

Nico69

Well-known member
Dec 28, 2019
194
188
4,770
Visit site
Can it really make that difference to digital? Is speed important? Is speed affected by cable type? For instance in my office at work we are on a Gigabit switch with superfast speeds on the Cable is only Cat5 not even 5e and it's a long wat from the local switch. Maybe 50m.

Screenshot 2023-03-17 105055.jpg
 

Friesiansam

Well-known member
Can it really make that difference to digital? Is speed important? Is speed affected by cable type? For instance in my office at work we are on a Gigabit switch with superfast speeds on the Cable is only Cat5 not even 5e and it's a long wat from the local switch. Maybe 50m.

View attachment 4523
I don’t believe it does make a difference. If it does make a difference, why does Qobuz sound good on my headphone amp, via the pc and the internet?
 

Vincent Kars

Well-known member
Is speed affected by cable type?
No, exactly the reverse.
The higher the bandwidth, the higher the requirements for the cable becomes.
Most streamers do have a 100 Mbs port so even Cat 5 will do but nothing wrong of course with 5e.

Beware of network cables with metal connectors. If the shield is attached to both, this can cause a earth loop.
 
  • Like
Reactions: WayneKerr

abacus

Well-known member
Even with my failing hearing I can hear a clear difference between some Ethernet cables. Whether it’s the quality of the wire, the connectors or the shielding, they can and very often do sound different to one another. I have just replaced a 8m cat8 uGreen cable (which sounded very good) with something that sounds better but cost about 10 times the price. Generally I have been unimpressed with cheaper “audio” ethernet cables and swopped them out for a uGreen cat8 loom and used the extra money from selling my “audio” cables to buy something else.
Then, one day, my brother in law came around with a super high end Ethernet cable that had been universally very well reviewed. It “sounded” very good indeed but not as good enough to make me want to buy one, However, it did set me off on trying to achieve a cost effective upgrade that would go someway towards recreating the achievable audio improvements.
I like the sound of silver plated copper and pure silver conductors in audio connectors, so that’s what I looked for. I like value for money and I don’t mind getting my hands dirty, so I thought I might try making a cable. I noticed most high end Ethernet cables use Telegartner connectors. My initial thoughts were simply to try replacing the connectors on my uGreen cables with Telegartner connections and see if that improved things, but as the connectors cost more than the cables I thought I would go the whole hog. and change the cable as well. I then found DH Labs reunion cable was available off the reel., it was cat8 (best shielding spec) silver plated copper cable. I started with a making a single length, swopping out the final length of uGreen cable to my streamer and was initially not blown away by the very slight change. I left my system running over the next couple of days and presto! more detail and a generally smoother and less harsh sound quality. I have slowly changed the three lengths of cable between my NAS and streamer and each increment was a significant improvement. The final link was the long 8 m length. Not great value for money, but every little bit helps.
Making the cable was easy and I already had the tools needed. (stanley knife, wire stripper, mole grips, small adjustable spanner).
I would say, without trying to sound elitist, that unless you have a high end and super revealing system you will hear so little difference (if any) that you should stick with a good quality standard cat8 Ethernet cable and spend your money on other things Like a bottle of wine or putting the heating on.

I presume all the differences you heard you were using a double blind test and not just swapping cables and listening. (if not your results cannot be verified and so are meaningless to anyone but yourself)
BTW: No one anywhere on the planet has yet been able to provide verifiable evidence that ethernet cable affect the sound, so you could be the first. (Assuming your claims are verifiable)

Bill
 

AndrewF

Well-known member
Feb 3, 2023
70
54
620
Visit site
Wait- this is a serious discussion? Ethernet cables?

(although, now that I think about it, my writing at work does occasionally seem to lack focus, depth of field and warmth....)
 
Last edited:
  • Like
Reactions: clanking

clanking

Well-known member
Feb 6, 2023
28
13
545
Visit site
If your cable isn't faulty then there is literally no difference between Ethernet cables that could possibly in any way affect sound quality.

Get a cheap switch that shows you stats on packet loss as that's the only thing that could impact anything and it's extremely unlikely in a residential setting unless the cable is badly routed next to some awful sources of interference.

If you've bought anything more than cat6a you've been conned, and cat6a is only really useful if you are running 10gbit networks. There is no such thing as cat 7/8 as a standard. Those that are marketed as such are intended for use in data centers where they bundle up hundreds of Ethernet cables in big runs between racks. Not for use in a house.

Even if there was packet loss it'd result in pauses, skips and broken playback, not some sort of 'made the treble a bit wooly and the bass went flubby' nonsense. The signals sent over digital networks are binary and the reason binary is used is because it can very easily be read and sent unambiguously with unsophisticated hardware over distance. A £1 cable into a £10 switch can send data just as well as a £30 cable into a £100,000 switch. It's digital, it'll be bit perfect, that's the point of it.

This is why things such as audiophile switches are snake oil and if you bought one you've been taken for a ride. Go to a large data center and you'll not see such things and they're the switches that are sending the streamed data to you.

Fwiw, I buy a cheap 50m reel of cable and add the rj45 connectors myself. All my switches are fully managed and I have no packet loss.
 

TRENDING THREADS

Latest posts