Thread: Anyone interested in some Ethernet cable evaluation info?

Hey guys,
I've been evaluating dfiferent Ethernet cables from my downstream FMC to my streamer for the last week or so to determine which Ethernet cable I want to live with in my system in the long term. The cables I've been evaluating by subjective listening tests are Audioquest Cinnamon, AQ Vodka and Wireworld Starlight Cat 8, and in a week or so, Supra Cat8.

I posted my impressions on Steve Hoffman forums, which, quite predictably, generated a very large degree of flak/noise/BS/insults from all the "cable objectivists" regarding my comments about the audio qualities of different Ethernet cables.

So, in this case, I'm asking FIRST if folks are sincerely interested in my strictly subjective listening impressions of using Ethernet cables in a streaming setup. NOTE: This will NOT be a scientific-based comparison, ABX-testing, randomized double-blind study with placebos, or anything of that type. There will be no math or equations characterizing VP/frequency or inductance down a cable at RF. This is just me listenin' to some Ethernet cables. Nothing more.

So. If there are folks who are truly interested in this topic, I'm happy to post my notes and impressions.

HOWEVER: if there are folks who simply cannot resist themselves and feel compelled to crap all over another "cable thread", particularly a thread about Ethernet cables (YOIKS! likely the most contentious class of cables since the Dark Days of the Power Cable Wars ), then I'll keep my findings to myself, as I simply don't want to yet again wade through all the cr*p/noise/flak I had to in my thread on SHF (I literally had guys insulting me personally about my physical hearing abilities).

I'm fairly sure some of you will be interested in my findings, but I'm don't want to start a thread that will decay into yet another "cable war" or generate a lot of "noise" in what is a friendly forum.

So...you can guys can let me know. Thanks.

the 'ole Puma Cat

Please do, it is always interesting and from my perspective, very valuable information. Thank you for sharing!
Best, GN

Let's hear your subjective test info bro
We appreciate your effort


Sent from my iPhone using Tapatalk

When I bought my Lumin streamer, the sales person told me that an ethernet cable is an ethernet cable.

I'd love to know if a better cable would improve the SQ. I haven't tested any myself, but I'm certainly open to the idea of one being better than another.

Originally Posted by Mike

Steve, you bought from the wrong dealer!


Sent from my iPhone using Tapatalk

LOL, Mike.

Based on what I learn from this thread, I may have to set him straight.

Originally Posted by Mike

Yes! Fire away.

Will do, Mike.

BTW, I will be calling you today regarding a question with respect to the Keces P3.

Another vote for your impressions!

Originally Posted by Huite

Thorough shielded (and good) cables everywhere I would say!

Actually, no, shielding is not good everywhere. Its appropriate for ICs and digital coax cables, but not desirable for other types of cables, most notably speaker cables.

Here is some information from Galen Gareis, who is a professional cable R&D principal design engineer at Belden. From his design brief for Iconoclast cables (manufactured by Belden):
Shield material and design considerations:
"I kept this topic here on purpose. Some may already know that low impedance cables signal levels
negate the need for a shield. And that’s a good thing because a shield over a speaker cable is darn
near ALWAYS a bad thing for two reasons;

1. A shield will always increase capacitance of the cable. The question is how much.
2. To mitigate the capacitance increase, the shield must be moved significantly AWAY from the
core polarities, increasing the size of the cable.

He continues: "Shields are ONLY beneficial if the environment demands them. Shields inhibit the performance of cable in most cases. Coaxial cables being an exception as the shield defines the cable’s natural IMPEDANCE. The ground plane proximity and uniformity are vitally important with short wavelength RF cables. Coaxial cables do just that. Audio is not RF, and these shields are more FUD devices than actual benefits, especially in speaker cables that have signals orders of magnitude over the background noise. Incidentally, the woven pattern in ICONOCLAST has a built-in immunity to RF not that that RF immunity is evident in the use of the cable.

View a SHIELD as a rain coat; great if you have water flying around but a major hindrance if you
don’t. Audio seldom needs shielding on low impedance cables and here is why;

Magnetic fields decay rapidly with distance; ratio of 1/x^3. The best defense is to MOVE the low frequency electromagnetic cables away from one another. The foil and even braid shields are higher frequency shields that are ineffective at much below 1 MHz. Magnetic fields lines need low permeability shield material (something a magnet will stick to) to route flux lines away from sensitive
devices. A faraday cage is an example you can put something into to do this. Low permeability metallic shields are a pain to use (stiff and heavy). DISTANCE is the best remedy. For EMI and RFI, the foil and braid shields used on Interconnect cable will be fine for RFI
ELECTRIC field issues, but NOT 20Hz-20KHz magnetic fields. "

If you're interested in reading the entirety of the design brief, including the equations that underpin Galen's remarks, let me know and I'll add the URL.

Originally Posted by joeinid

+1

Absolutely interested.

Hi Joe! Good to be here (again...)

Originally Posted by kiwi_1282001

Yes Puma Cat, so long as the only variable is a change in a single Ethernet cable, I would be most interested to read your findings. This is not "Steve Hoffman forums".

Hi Kiwi,
That is exactly the only variable that will be changed, a single Ethernet cable (from my downstream FMC to my Sonore microRendu. Look for my notes on the configuration in the introductory/context post

Originally Posted by aKnyght

very interested...

as a baseline, you might add generic cat5e or cat6 (unshielded, twisted pairs) and cat7 (shielded, twisted pairs) cables to the mix. also, comparing these two will tell you about the impact of shielding on this link in your system... there have been some reporting that shielded cables introduced a ground link through the shield which is, obviously, detrimental to SQ.

looking forward to your findings, regardless of the sample set you choose.

Yep, I can do that; I have some generic Cat 5 and Cat6 cable around here I can plug in. I also have some Belden Cat6e cable, but I am not sure I want to plug it in; the last time I did so, the sound was pretty ghastly. Maybe it needs some burning in...I'll see if I can set something up to burn the Belden over the next week or so...

Originally Posted by UltraFast69

It would be interesting to hear between the two for audio.

I would never think not to choose shielded Ethernet cables.


Sent from my iPhone using Tapatalk Pro

Depends: I've heard pros and cons from different folks on shielding for Ethernet cables. Just as a general rule, sheilding is not always desirable, and for some applications, not the way to go. See my comments based on Galen Gareis' views above.

Originally Posted by Tubes

When I bought my Lumin streamer, the sales person told me that an ethernet cable is an ethernet cable.

I'd love to know if a better cable would improve the SQ. I haven't tested any myself, but I'm certainly open to the idea of one being better than another.

Well, respectfully, I don't agree with that sales person. I wish that was the case, actually, because I then wouldn't have to buy an "audiophile-grade" Ethernet cable.

But, it's not.

So, as an foundation for this evaluation here is the configuration I am running: Ethernet->Fiber->Ethernet->Streamer config for very specific reasons I won't belabor here.

In my small townhome, a long run of Ethernet to a GigaFoil IV and then Ethernet to my streamer is simply not practical, so my run of optical fiber is ~ 7 meters of Tripp-Lite MMC SC/SC fiber connected by two 850nM TP-Link fiber media convertors (FMCs); this saves on requiring a long run of Ethernet cable, and mitigates/attenuates the problems concomitant with a long run of Ethernet, e.g. EMI, RF, lack of galvanic isolation, etc.

The streamer is presently a Sonore microRendu powered by an Uptone Audio LPS-1 linear power supply which in turn is powered using a Shunyata Venom 14 power cord plugged into a Shunyata Triton distributor. The upstream FMC is powered by a Jameco 9V low noise regulated linear power supply. The downstream FMC (the one that connects to the streamer) is powered by a 9V iFi iPower LPS plugged in my Shunyata Tirton. So, the power going to the Sonore is pretty darn quiet, but not as quiet as it could be with a Keces P3 (Mike and I discussed that on the phone today).

The USB cable going from the Sonore is a Shunyata Alpha USB cable to my Schiit Gungnir Gen 5 USB multibit DAC. The Schiit Gumby is powered with a Shunyata Black Mamba CX PC, also plugged into the Shunyata Triton. Preamp is a Conrad-Johnson CT-5 and power amp is a C-J LP70S, all plugged into the Triton with Shunyata Black Mamba CX (HC/CX for the power amp) or Zitron Cobra PCs. Speakers are Dynaudio Contour S3.4s with Esotar 2 tweeters and Shunyata Venom speaker cables.

So, needless to say, this is a pretty dang quiet, transparent and revealing digital front-end and system, overall. It also sounds very lovely, sweet, natural, analog, and musical in a very nice way. (Note: The Shunyata Alpha USB took this system to another level I simply have not heard on a digital front-end before).

So...that's the system and configuration...

Originally Posted by Puma Cat

Well, respectfully, I don't agree with that sales person. I wish that was the case, actually, because I then wouldn't have to buy an "audiophile-grade" Ethernet cable.

But, it's not.

The whole ethernet/streaming thing is too new to me for me to have an opinion either way so I'll be anxiously awaiting your results.

Originally Posted by Tubes

The whole ethernet/streaming thing is too new to me for me to have an opinion either way so I'll be anxiously awaiting your results.

Thanks! Its fairly new to me too, though I have been using my Sonore microRendu since the end of 2016. The BIG improvement came when I moved the Mac Mini music server out of the audio rack and put it at the other end of the house.

So, here are the cables I'll be describing: Audioquest Cinnamon, 0.75 M, Wireworld Starlight Cat8 1.5M, and Audioquest Vodka, 0.75M, Supra Cat8 1M, a $7 7M run of Amazon Cat7 cable, and a generic Cat 6 unbranded cable I've had around the house for some years now. The WW Starlight Cat8 and AQ Vodka are rented in from The Cable Company. I own the Audioquest Cinnamon, and bought and installed back at the end of 2016 when I installed the Sonore microRendu. It will serve as my baseline reference point. The Supra Cat8 are new and are being "burned in"; I'll comment on these in a week or so vs. the baseline reference, Cinnamon.

My Audioquest Cinnamon, which is well burned-in is a very nice-sounding Ethernet cable; reasonably spacious, natural and neutral with a hint of warmth in overall character. There's a nicely defined but somewhat smallish soundstage and good but not exceptional imaging. The overall musical presentation is crisp, clean and quiet, and the musical presentation quite natural-sounding, just on the warm side of neutral. The soundstage does not extend past the sides of the speakers. By contrast, the $7 Amazon Cat7 cable I was using before I installed the run of fiber sounds muffled, veiled, closed-in, dark, and most notably, quite rolled off on the top end, giving a very closed-in sounding top end. Cinnamon is more extended on the top end. Cinnamon is not quite as extended or silky as the AQ Vodka (more on that later), but its very similar overall in character and, at $80 for 0.75M, what I consider to be quite a good value for money.

Originally Posted by Huite

[FONT="]I use speaker cables from the British Chord Company. Cables from their top range "Chord Music". [/FONT][FONT="]https://www.chord.co.uk/product/chordmusic-speaker-cable/
[/FONT]These cables are shielded in 7 different ways. And believe me, the Chord Company is well-known and my speaker cables sound great. So, as we say over here, there are many ways that can lead you to Rome!
To return to ethernet cables. Here is a short video with Nigel Fynn who is technical director of te Chord Company. At the end he says some interesting things about ethernet cables. https://www.youtube.com/watch?v=gSgL9JlehjM

Knowing full well Chord's reputation for making excellent-sounding products, I'm sure they sound great. And thanks for sending the link to the video. I watched the interview with Nigel with interest. In particular, his comment about changing "the spacing between the conductor and the shield" is fully consisent with point that Galen Gareis was making in his design brief for his speaker cables, specifically:

"Magnetic fields decay rapidly with distance; ratio of 1/x^3. The best defense is to MOVE the low frequency electromagnetic cables away from one another. The foil and even braid shields are higher frequency shields that are ineffective at much below 1 MHz. Magnetic fields lines need low permeability shield material (something a magnet will stick to) to route flux lines away from sensitive devices. A faraday cage is an example you can put something into to do this. Low permeability metallic shields are a pain to use (stiff and heavy). DISTANCE is the best remedy."

This is exactly what Nigel said that Chord did with the "Signature Reference".

They increased the spacing between the conductor and the shield because it makes the Chord cable perform and sound better for exactly the reasons that Garies cites above.

Nigel also commented that the new speaker cables are significantly less stiff than the older model, but added, "...relatively speaking.." This is also consistent the Galen's comment that low permeability shields are a pain to use, stiff and heavy.

I'm sure the Chord Signature Reference sound great; Chord is a solid engineering-based audio company with excellent credentials. But, respectfully, I'm not convinced that because Chord does it that it is a cable design paradigm or "central dogma." Digital coax, sure. Speaker cables, no. Personally, I'm an advocate for "fit for purpose" engineering. I don't see an advantage for engineering something just because you can. This is what Howard Hughes got caught up with with the Spruce Goose. And, particularly if it adds unnecessary complexity to a design embodiment, or, most importantly, if it adds undesired effects that then have to be compensated for.

I'll give you example. For years now, Audioquest has been installing Dielectric Bias System (DBS) units to their cables. They actually work, and reduce the time for burn-in. But, as they are battery powered and output 72 volts, it turns out that the DBS units also output RF. So, now, AQ had to modify the DBS units to a new engineering specification to include an "RF trap" to protect the cable from the RF that the DBS field elements themselves put out.

*-unrelated side note: Why do almost all audiophile product companies use terms like "Reference, Statement, Signature, Platinum, or Reference Statement or Signature Statement or Reference Signature, etc.? I find it amusing, but I digress.

Originally Posted by mep

We are on page 4 of this thread. When does the cable evaluations start?

They've already started. My impressions of AQ Cinnamon are above.

Okay, here are my impressions of WW Starlight Cat8:

WW Starlight Cat8 cable back in the system after the AQ Cinnamon. Its a nice-sounding cable, on the whole, its most notable attribute is that its very quiet, and compared to AQ Cinnamon is that the soundstage on the Wireworld Starlight Cat8 is quite a bit wider than Cinnamon; it extends well past the sides of the speakers; Cinnamon's soundstage does not extend past the sides of the speakers. And, there is more space around each instrument and voice in the stereo image. The overall soundstage is on the two-dimensional side in that its not particularly deep front-to-back. The musical presentation is very neutral, but on the lean, cool side of neutral than the warmish-side of neutral. Vocals and instrumentals are clean and well-defined, but what I would classify as somewhat spare and thin in presentation and body. String bass definition and articulation is excellent, and a just a touch dry and lean. Some sibilants can be somewhat hot and/or bright, depending on the recording.

The Audioquest Vodka is fuller and richer-sounding than Wireworld Starlight Cat 8, but the soundstage does not extend as wide. The soundstage with the WW Starlight Cat 8 extends well past the speakers and the separation of instruments in the sound field is larger and more spacious, the Vodka just from the middle to the outer edges of my floor-standers and instruments are not as spatially-separated as much as they are on Starlight. The Vodka is harmonically richer and fuller-sounding with greater dynamic slam than Starlight Cat8 and with a deeper soundstage front-to-back, but sting bass not as tightly defined and articulated; its a a bit looser and flabbier than Starlight Cat8. Starlight 8 is just on the lean side of neutral, Vodka is just on the warm side of neutral. Overall, I prefer AQ Vodka to WW Starlight as it is more natural and “musical-sounding”.

Originally Posted by Puma Cat

Hey guys,
I've been evaluating dfiferent Ethernet cables from my downstream FMC to my streamer...

I'm embarrassed to say that I don't know what FMC stands for. I tried Googling it but somehow I don't think it stands for Federal Maritime Commission, in this context.

Originally Posted by mep

So if the bass from the Vodka is not as "tightly defined and articulated" and is "a bit loose and flabbier" than the Starlight Cat 8 cable, how can you possibly say that it is more natural and "musical sounding?" Loose and flabby bass has never sounded natural to me nor is it something I want to inject into my system.

Ah okay, sorry. So, for clarification and in the interest of accuracy, I didn't say the bass was "looser and flabbier", I said it was a "bit loose and flabbier." The distinction is important becaue the difference is relative. A "bit" means "a little". For string bass, like Ray Brown's bass on Cry Me a River on Soular Energy, its as not as tightly defined and articulated as Starlight. Part of this is because on the whole, Starlight is drier and leaner in overall character than Vodka. This particular characteristic brings slight advantage with respect to reproduction of the timbres of the string bass, specifically. But the dryness and leanness of Starlight provides downsides in other parts of the presentation, overall. Starlight is somewhat on the sterile-sounding side of things. Its nice, don't get me wrong, its just not as nice as Vodka, in my system (ask Mike, components and cables interact to produce different responses in different systems).

With respect to more natural and musical-sounding, I was referring to the overall musical presentation, taken as a whole. Overall, the attributes the Vodka has brings, in my system, a more engaging experience than Starlight Cat8.

Originally Posted by Tubes

I'm embarrassed to say that I don't know what FMC stands for. I tried Googling it but somehow I don't think it stands for Federal Maritime Commission, in this context.

FMC is a Fiber Media Convertor. The fiber media convertor accepts an Ethernet cable via an RJ45 jack. It also has an optical transceiver module that accepts a fiber cable. The FMC converts the analog electrial square waves passing down the Ethernet cable's metal (usually copper) conductors into light signals and sends them out via the FMC's optical transceiver output port.



Because the signals travelling down the fiber are light and not an electromagnetic square wave, fiber is impervious to the effects of EMI, RF, and other high-bandwidth noise and also provides galvanic isolation. This results in a notably quieter noise floor and improved sound quality.

I use two of them, one at the upstream end, and a downstream FMC, near the streamer, and a 7M run of Tripp-LIte fiber instead of a long run of very expensive audiophile-grade Ethernet cable. Instead, I only need a 0.75M length of audiophile-grade Ethernet cable going from the downstream FMC to the Sonore.

Example: A friend of mine on another forum recently bought a 10M run of WW Starlight Cat8. 10M of Starlight Cat8 costs $750. A 10M run of Platinum Starlight costs...$6250! My two TP-Link FMCs and a the run of fiber cost me...$66. Installing it resulted in a noticeable drop in the overall streaming digital front end's noise floor and a much more open, extended and airy top-end.

Originally Posted by Puma Cat

FMC is a Fiber Media Convertor. The fiber media convertor accepts an Ethernet cable via an RJ45 jack. It also has an optical transceiver module that accepts a fiber cable. The FMC converts the analog electrial square waves passing down the Ethernet cable's metal (usually copper) conductors into light signals and sends them out via the FMC's optical transceiver output port.



Because the signals travelling down the fiber are light and not an electromagnetic square wave, fiber is impervious to the effects of EMI, RF, and other high-bandwidth noise and also provides galvanic isolation. This results in a notably quieter noise floor and improved sound quality.

I use two of them, one at the upstream end, and a downstream FMC, near the streamer, and a 7M run of Tripp-LIte fiber instead of a long run of very expensive audiophile-grade Ethernet cable. Instead, I only need a 0.75M length of audiophile-grade Ethernet cable going from the downstream FMC to the Sonore.

Example: A friend of mine on another forum recently bought a 10M run of WW Starlight Cat8. 10M of Starlight Cat8 costs $750. A 10M run of Platinum Starlight costs...$6250! My two TP-Link FMCs and a the run of fiber cost me...$66. Installing it resulted in a noticeable drop in the overall streaming digital front end's noise floor and a much more open, extended and airy top-end.

Thank You Puma Cat for your thorough explanation.

Originally Posted by bart

Stephen, it would be great for some of us, ignorant beginners in streaming, if you provide your exact set-up (with brand names of all devices in the chain).
You seem to find a way of optimising your system without breaking the bank (too much).

Can do, Bart.

Mac Mini-> (originally generic, now Supra Cat8)1M Ethernet cable > Pace Router > (originally generic, now Supra Cat8) Ethernet -> "upstream" FMC -> 7 M (23 ft) optical fiber -> "downstream" FMC -> Ethernet cable ("audiophile grade" under evaluation)-> Sonore microRendu -> Shunyata Alpha USB -> Schiit Gungnir Multibit DAC > CJ CT-5 preamp.

The Mac Mini, which is in another part of the house, as far away from the audio rack as possible, is powered with a Shunyata Venom 14 with a C7 plug. I'll probably move my Hydra 4 in there to power it.

The Sonore microRendu is powered by an UpTone Audio UltraCap LPS-1 PS with some ridiculous amount of capacitance reserve (60 Farads). The Meanwell AC/DC transformer for the LPS-1 is powered with a Shunyata Venom 14 plugged into my Shunyata Triton distributor. The upstream FMC is powered by a low-noise Jameco 9V 500mA 4.5W regulated linear power supply (P/N GPU410900500WDOO). This power supply is plugged into a $39 Audioquest IEC>3US power-strippy thang that is connected to the wall receptacle with a Shunyata Diamondback power cord. The downstream FMC, which is near the audio rack, is powered by a 9V iFi iPower linear power supply plugged into the Shunyata Triton.

The two FMCs are TP-Link MC200CM SC/SC 850nM Multi-mode fiber media converters. The fiber is Tripp-Lite, 7 meters, 62.5/125 fiber patch cable SC/SC, P/N N306-07M.

My "reference" Ethernet cable from the downstream FMC to the Sonore microRendu is an Audioquest Cinnamon, 0.75M

Hope this helps...