Ethernet cables, Switches and Network related. Based on listening is it worth to buy more expensive ones? Share your listening experience.
Sep 7, 2021 at 9:10 AM Post #481 of 639

The Jester

100+ Head-Fier
Joined
Jun 21, 2020
Posts
441
Likes
354
Location
Australia
Totally dependant on the DAC involved for sure, some like the Chord DAC’s have had very good results with jitter measurements, but there are so many variables out there in cables, DACs, filters, reclockers etc all we can do is try it and see, going by recommendations of other systems will get you so far, after that it’s down to YMMV …
 
Sep 7, 2021 at 9:54 AM Post #483 of 639

griff500

500+ Head-Fier
Joined
Oct 19, 2020
Posts
609
Likes
796
Location
UK
Yes they do and yes they do.

I feel the discussion has progressed nicely.
 
Sep 7, 2021 at 10:08 AM Post #485 of 639

griff500

500+ Head-Fier
Joined
Oct 19, 2020
Posts
609
Likes
796
Location
UK
Presumably based on your practical experience rather than just theoretical.
 
Sep 7, 2021 at 5:10 PM Post #486 of 639

Dawnrazor

500+ Head-Fier
Joined
Dec 22, 2017
Posts
539
Likes
147
Location
FLorida
All happens at the final conversion, if the DAC has a good clock, power supply, galvanic isolation etc it should be fine …
yep. Do you know of a dac that is totally immune?
 
Sep 7, 2021 at 5:28 PM Post #487 of 639

The Jester

100+ Head-Fier
Joined
Jun 21, 2020
Posts
441
Likes
354
Location
Australia
Totally is a big word,
I don’t swap DACs that often, used an Eximus DP1 with its then new Xmos USB input for over 5 years with for me great results,
recent upgrade to a Chord MScaler/Qutest is better again,
Have a look at some reviewed measurements of Chord DACs in particular the jitter levels …
 
Sep 7, 2021 at 9:37 PM Post #488 of 639

Dawnrazor

500+ Head-Fier
Joined
Dec 22, 2017
Posts
539
Likes
147
Location
FLorida
Totally is a big word,
I don’t swap DACs that often, used an Eximus DP1 with its then new Xmos USB input for over 5 years with for me great results,
recent upgrade to a Chord MScaler/Qutest is better again,
Have a look at some reviewed measurements of Chord DACs in particular the jitter levels …
I was looking at the Qutest but I saw a review where a guy owned it and thought he was done with dacs. But he ended up with another dac and while on swapping the Qutest was better, he kept coming back to the other dac and eventually sold the Qutest. I am betting my house that that dac is horrible in the measurement departments..... I bought that dac and really dig it.
 
Sep 8, 2021 at 2:43 AM Post #489 of 639

The Jester

100+ Head-Fier
Joined
Jun 21, 2020
Posts
441
Likes
354
Location
Australia
I was looking at the Qutest but I saw a review where a guy owned it and thought he was done with dacs. But he ended up with another dac and while on swapping the Qutest was better, he kept coming back to the other dac and eventually sold the Qutest. I am betting my house that that dac is horrible in the measurement departments..... I bought that dac and really dig it.
Depends On the rest of the system and our own preferences, too much can be made of measurements … S/N ratio without any measurement of noise floor modulation, THD figures that don’t highlight differences between smoother sounding 2nd harmonics and less pleasant 3rd harmonics,
with Jitter levels though less is usually better …
 
Sep 10, 2021 at 11:21 AM Post #490 of 639

cpurdy

Head-Fier
Joined
Aug 18, 2019
Posts
78
Likes
84
Location
Boston MA USA
ethernet cables effect on the audio, really is nothing to do with packets or changing the data in anyway. You are right ethernet is a robust way for delivering data, but it wasnt designed with audio in mind. The biggest issue effecting audio performance is the transmission of electronic (main RF) noise, picked up from around your network and transmitted via the ethernet cable acting if you will, like an antenna. This noise then piggy backs the data and enters the streamer which is highly sensitive to the effects noise brings, ie vibration and jitter which in turn causes an adverse effect when it comes to the DA conversion later in the chain. For the majority of applications ethernet works seamlessly, but in audio the physical connection and the noise which can be transmitted does effect the way your streamer and DAC perform and consequently the sound we hear.
This ^ is just made up bullsh1t.

To get sufficient noise on the cable to impact electronics on either end, you'd be completely destroying the data going on the wire. Unless you're placing a running a vacuum cleaner and hair drier on top of the ethernet cable, you're not experiencing this.

And if you have a DAC with a crappy enough design (a single packet sized buffer?) that it could be effected by a dropped (and then automatically re-sent) ethernet packet, then seriously, W.T.F?!?

After some time I replaced the upgraded ethernet cable with the original one and the sound quality was worse - immediately and easily noticeable. I really do not need to confront anything about my purchasing decisions and it is not incumbent on me to prove anything to you. Instead of that, I'll enjoy my music, which sounds a lot better with this cable.

Expectation bias is a two-way street. My firm expectation was that an ethernet cable would make no difference to a digital signal, but I thought I would give it a go anyway, with the ability to return the cable. I was wrong. Of course, you will continue to believe that everyone who says they hear a difference is wrong and it won't raise any curiosity whatsoever in your enquiring scientific mind...
If it wasn't a reproducible, double-blind test, then it's just the placebo effect.

There's nothing wrong with the placebo effect. It's an awesome effect. If it makes your enjoyment of audio better, then by all means, turn up the placebo! I personally rely on the placebo effect all of the time, and have no shame in doing so. I spend good money on AV placebo equipment. And I love it.

But I am careful not to mislead others about the difference between reality and placebo.

It seems somewhat arrogant for you to think people will find your posts valuable, and particularly when they are purely theoretical.
We seem to have different understandings of the word "theoretical".
1631285828885.png


On the one hand, we have actual, measured, tested, real-world proven, fifty years worth of proof that basic datagram packets over Ethernet could experience an error rate of statistical zero (not actual zero, but close enough that you are unlikely to experience a single such error in your lifetime). And that's before you add on any high level protocols that can easily provide an actual zero error rate.

On the other hand, we have people making ridiculous, anti-reality claims, backed up solely by "I may sell these expensive cables to make money, but trust me, I can hear the difference!"

The galvanic isolation most be to some extent in some limits. I think the noise we are talking about that effects audio dont go over the threshold where it can make stuff malfunction. The data is just electric voltage signals to my understanding and that passes on and probebly with it noise of some sort. Why would noise in ethernet wire cables be something magical dead set vaIue, i belive in gradations of noise. Does anything exist without gradations? Maybe that would be the odd thing "Ethernet the stop dead halt of everything Except magical data".
I guess you need to cut the wired ethernet cable to be compleatly isolated but no data (electrical signals) will pass then neither.

If it was galvanicaly isolated completely would this occour then? https://www.google.com/amp/s/www.bbc.com/news/uk-wales-54239180.amp
Yes. Digital data is "just electric [sic] voltage signals". Sure, a few trillion dollars has been made by figuring out how to correctly deal with those "just electric [sic] voltage signals". Minor little start-ups like Intel, ARM, Asus, AMD, Fairchild, Apple, AT&T, Acer, IBM, TSMC, Samsung, Honeywell, Western Digital, SK Hynix, ...

Dealing with those signals let us go to the moon. Allowed us to decode the human genome. Let's us fairly accurately predict weather. Take pictures of distant galaxies. Carry an always-connected combination GPS + computer + telephony device in your pocket. Little things like that.

As an audio equipment maker, you'd have to be pretty stupid to route noise from the PHY of the NIC into your amplifier, but I guess that is possible to achieve, if you wanted to. However, I'd suggest avoiding audio equipment from anyone who designs their equipment to route noise from the PHY of the NIC into the analog output from the DAC.

And the article that you linked to has nothing to do with Ethernet.
 
Last edited:
Sep 10, 2021 at 12:02 PM Post #491 of 639

teknorob23

Member of the Trade: Network Acoustics
Joined
Feb 28, 2018
Posts
1,607
Likes
1,814
Location
UK
This ^ is just made up bullsh1t.

To get sufficient noise on the cable to impact electronics on either end, you'd be completely destroying the data going on the wire. Unless you're placing a running a vacuum cleaner and hair drier on top of the ethernet cable, you're not experiencing this.

And if you have a DAC with a crappy enough design (a single packet sized buffer?) that it could be effected by a dropped (and then automatically re-sent) ethernet packet, then seriously, W.T.F?!?


If it wasn't a reproducible, double-blind test, then it's just the placebo effect.

There's nothing wrong with the placebo effect. It's an awesome effect. If it makes your enjoyment of audio better, then by all means, turn up the placebo! I personally rely on the placebo effect all of the time, and have no shame in doing so. I spend good money on AV placebo equipment. And I love it.

But I am careful not to mislead others about the difference between reality and placebo.


We seem to have different understandings of the word "theoretical".
1631285828885.png

On the one hand, we have actual, measured, tested, real-world proven, fifty years worth of proof that basic datagram packets over Ethernet could experience an error rate of statistical zero (not actual zero, but close enough that you are unlikely to experience a single such error in your lifetime). And that's before you add on any high level protocols that can easily provide an actual zero error rate.

On the other hand, we have people making ridiculous, anti-reality claims, backed up solely by "I may sell these expensive cables to make money, but trust me, I can hear the difference!"


Yes. Digital data is "just electric [sic] voltage signals". Sure, a few trillion dollars has been made by figuring out how to correctly deal with those "just electric [sic] voltage signals". Minor little start-ups like Intel, ARM, Asus, AMD, Fairchild, Apple, AT&T, Acer, IBM, TSMC, Samsung, Honeywell, Western Digital, SK Hynix, ...

Dealing with those signals let us go to the moon. Allowed us to decode the human genome. Let's us fairly accurately predict weather. Take pictures of distant galaxies. Carry an always-connected combination GPS + computer + telephony device in your pocket. Little things like that.

As an audio equipment maker, you'd have to be pretty stupid to route noise from the PHY of the NIC into your amplifier, but I guess that is possible to achieve, if you wanted to. However, I'd suggest avoiding audio equipment from anyone who designs their equipment to route noise from the PHY of the NIC into the analog output from the DAC.

And the article that you linked to has nothing to do with Ethernet.

do you need a hug?
 
Sep 10, 2021 at 12:03 PM Post #492 of 639

griff500

500+ Head-Fier
Joined
Oct 19, 2020
Posts
609
Likes
796
Location
UK
If it wasn't a reproducible, double-blind test, then it's just the placebo effect.
I think you mean that it could be placebo.

Stating that it is placebo if it wasn't established through 'double blind' testing as if it's a fact simply shows your bias and your conclusion is flawed as you disregard the possibility that what was heard was real.
We seem to have different understandings of the word "theoretical".
Perhaps. Memes are not a great way of trying to make a point in an adult manner. The putdowns and condescension aren't a good look either.

My meaning is that if you have not tried something (practical experience) then your opinion is theoretical.

You bang on about data integrity when nobody is debating that. Yet again, it's not just about that. Perhaps you are in the group that thinks if something measures well then it must sound good. Not everything works exactly as expected - theory and reality do not always match each other, as any decent electronics engineer would tell you.

I think neither side are going to persuade the other and that's fine. You believe it makes no difference and most who have tried it found that they heard a difference. You believe they are all imagining it but they don't and they are happy with their systems.

Perhaps live and let live rather than trying to put people down?
 
Sep 10, 2021 at 12:04 PM Post #493 of 639

griff500

500+ Head-Fier
Joined
Oct 19, 2020
Posts
609
Likes
796
Location
UK
Sep 10, 2021 at 2:13 PM Post #494 of 639

audiobomber

500+ Head-Fier
Joined
Jul 22, 2018
Posts
521
Likes
402
Location
Sudbury, ON, Canada
IME, network tweaks can make worthwhile improvements in sound quality. I've only read the first and last few pages of this long thread, but here are my efforts:
- Replaced the SMPS supplied with the cable modem from my internet provider. This 12V 2A wall wart was corrupting the sound. I suspect it was defective, but if so there was no impact on network function. Replacing it with an iFi iPower SMPS removed sibilance and cleaned up the sound in all of my audio systems, even the two systems that play from the NAS, and not the internet.
- Added a Furman power conditioner for the network distribution gear
- Replaced the router wall wart with a Zero-Zone 12V 3.3A LPS
- Connected my network devices to a TP-Link switch instead of the router
- Added an iFi iPower X to the switch

I had tried a Welborne Labs LPS on the ethernet switch, but it was actually a step backwards vs. the supplied wall wart. I had read internet posts by a couple of different manufacturers, who claimed that some purely digital devices sound better with a switch mode supply. I purchased a new 9V iPower X and tried it vs. the Welborne LPS on my sMS-200 streamer and found them to sound the same. On the ethernet switch however, the iPower X was quite superior, so I left it there.

All of the above made small improvements that are nevertheless important to me as an audiophile.

My main audio system is connected to the network via a TP-Link extender, configured as a "Wi-Fi to Ethernet" bridge, with broadcast radio turned off. This was noticeably better than using the built-in Wi-Fi receiver in my exaSound PlayPoint streamer (Wi-Fi radios are electronically noisy, best kept away from the DAC). I replaced the SMPS on the extender and streamer with Zero-Zone linear power supplies. The LPS made an almost imperceptible difference on the PlayPoint, even though it is a higher spec 2020 S.S. version, but I like that I have no switch mode supplies in my main system.

The final tweak was ethernet cables. I worked at optimizing my cables for months, trying various Amazon generic CAT5e, 6, 6a, 7 cables and CAT8 cables from Supra and Yauhody. All sounded different, and significantly enough that they can improve or spoil the sound of my main system (significant but not as critical in my other three systems). Three cables stood out from the pack. I found the Supra CAT8 to be ideal tonally with great soundstage, but it softened initial transients in an unacceptable way. Tera Grand CAT7 had brighter highs and a more forward presentation, with fantastic bass definition. A little too forward for my system and preferences. Yauhody had a warm, organic sound, with a great soundstage that moved back a couple of steps. It was a tiny bit lacking in transparency.

One thing to be very careful about is that some CAT6a, some CAT7 and all CAT8 cables have a shield that is attached at both ends to the connectors. Depending on construction of the pieces on either end, this can result in ground current flowing from sender to receiver, defeating the inherent galvanic isolation normally provided by ethernet cable design. I tested for electrical isolation with an ohmmeter. The only place this was an issue in my system was from NAS to ethernet switch. This is where a LAN Isolator really matters, so I added an inexpensive LNF-C8G isolator from eBay in that position.

I have three ethernet cables in the network for my main system. People say only the final ethernet cable matters, but that is definitely not my experience. They all matter, but the final cable is most prominent. If I had to choose only one ethernet cable, it would be Yauhody CAT8, but no single brand I tried made me perfectly happy, so I tried mix and match combos. My final configuration is:
NAS -> Tera Grand CAT7 -> LNF-C8G LAN isolator -> Ethernet Switch -> Tera Grand CAT7 -> Router -> Wi-Fi to Ethernet bridge -> Yauhody CAT8 -> PlayPoint streamer.

Experimentation was very helpful. I think buying a bunch of basic cables and trialing them makes more sense than buying a single expensive audiophile cable, which may or may not suit your system. YMMV.
My audio systems in more detail here: https://audiophilestyle.com/profile/4137-audiobomber/?tab=field_core_pfield_3
 
Last edited:
Sep 10, 2021 at 5:39 PM Post #495 of 639

bluenight

500+ Head-Fier
Joined
Mar 31, 2017
Posts
535
Likes
132
Location
Sweden
You bang on about data integrity when nobody is debating that. Yet again, it's not just about that. Perhaps you are in the group that thinks if something measures well then it must sound good. Not everything works exactly as expected - theory and reality do not always match each other, as any decent electronics engineer would tell you.
I think your right. This is what ASR measures right.
 

Users who are viewing this thread

Top