Best audiophile switch

[NigelB]

This stuff is all over the internet and consistently described. I also discussed with Galen Gareis of Belden/BJC. They use 5c plastic ethernet connectors because they are the best, not because they cost 5c.

I have two very long runs and I use AudioQuest Pearl CAT6 in parallel with fibre. I like this cable because the conductors seem slightly thicker, which makes it a lot easier for people like me to use a crimping tool.

Plastic connectors are great for unshielded cables of course.

 

[NigelB]

There seems to be an audio mythology about music being extremely demanding on the network infrastructure. My understanding is that it is quite the opposite - it is generally a very low level signal. What is to be avoided is electrical noise getting into your streamer, whether through the shield, EMI or power supply.

Absolutely this.

Some folk seem to be confused by eg Cat 8 which is able to handle huge amounts of data precisely because of the way its high twist rates and excellent shielding mimimise the amount of extraneous noise and in-cable crosstalk; but we audiophiles are more interested in the side-effect (noise reduction/rejection) than the speed they are capable of.

Most of us continue to use switches capable of handling no more than 1Gbps, sometimes 100Mbps. Data bandwidth/speed in the cables we connect to these is hardly an issue.

 

[NigelB]

Oh dear, so much misinformation being put out there!

The tiny transformer magnetics at EVERY RJ45 Ethernet port in the world DO provide galvanic isolation. (Though some of those magnetics are better than others; most have 6~8 cores, the magnetics in EtherREGEN have 12 cores per port.) These transformers will block DC and low-source-impedance leakage. However, they DO NOT block high-source-impedance AC leakage.
Optical connections of course do block all leakage, but the circuitry in optical SFP transceivers generate their own noise at both ends and some amount of jitter (however small).


[As a footnote aside, since this is not a promotional forum, the--still 100% unique--advantage of the EtherREGEN (and why we have 'A' and 'B' side ports) is our active-differential isolation "moat". Literally high-speed differential digital isolation chips (expensive) sitting across a gap in the PCB.
Data--in SGMII format for Gen2 s we can offer 'B'-side Gigabit in Gen2 (isolators these fast were not available when back in 2018)--flows across this "moat", and clocks in differential format flow back to the 'A' side across another isolator. Following the isolators the data goes directly into our our secret 10GHz-capable ultra-low jitter (just 200 femtoseconds!) reclocking flip-flop. This is where the "magic" happens.]

Agreed: there is a lot of misinformation being put out there.

 

[Ground Control]

The tiny transformer magnetics at EVERY RJ45 Ethernet port in the world DO provide galvanic isolation. (Though some of those magnetics are better than others.

So what does this actually mean? Is isolation not binary? Does the effect diminish over time?

 

[kennyb123]

One thing I have always appreciated about Ken's system across its many evolutions (a handful of which I've been privy to) is that his focus on timing and precision makes it an excellent lens to experience music through. When you plug something new into his system, it's very 'what you hear is what you get' -- not that it's perfect (no system is), but that the relative strengths of whatever gear or cabling you insert are evident in some form. In other words, it doesn't take effort to hear a difference when there's a difference to be heard.

Returning to Jeremy's excellent post as there are a few comments I wanted to make.

Appreciate the kind words about my system. I think the room treatments really help to make it easy to notice differences. Those ASC tube traps and Stillpoints Aperture panels made a huge difference.

It was pleasing to me (hello, bias) that my ~$2300 Synergistic Enternet Switch UEF, when powered with its own Foundation SX 12 power cord, delivered to my ears ~85% of what the ~$4200 Network Acoustics Tempus switch did when it was powered with a Synergistic Venom cable.

It was interesting to hear how the Synergistic switch evolved right in front of us. This easily proved that it's worth the effort to try to get the best out of the gear we already purchased

Ken had borrowed these plugs from me before and he didn't find that they made any tangible difference in the etherREGEN. Certainly not a $300/ea difference.

I have been a fan of Synergistic's products that utilize UEF to reduce noise. I have a number of products that feature this (fuses, AC outlet, outlet cover, ECTs, Carbon Tuning discs) and they each delivered a benefit that justified what I paid for them. Thus, it really surprised me that these doohickeys didn't seem to do much when I tried them.

But that last ~15% is the difference between "this sounds good" and "this sounds real". And it's the 15% we audiophiles chase incessantly. It almost always lives on the "other end" of the price/performance curve, after the knee, climbing that asymptote to the sky where exponentially more money is required to realize each little step along the path of sonic bliss...

This is an excellent point. I do hope though that more can be achieved at a lower price as the switch marketplace matures. I think three out of the four switches were first generation products, with the Tempus possibly being a second generation as @keithc mentioned in passing that NA had a previous switch with a different name.

I'm not including the Innuos PhoenixNET in my ranking because it wasn't given the proper settling time for its clock (24h)

Another excellent point. I probably should have done the same. Keith spent more time with the PhoenixNET, so I look forward to reading his review once it's published.

was how good the 6m QSA-Lanedri Gamma Infinity Ethernet cable (~$6000) sounded without any switch at all

It was really good - especially into the switches. That's gonna cost me.

We all agreed that there was clearly more noise there and perhaps the most vital function of the switches was to filter out the noise. The degree to which they filter it out, and the adeptness with which they refrain from imparting their own, seems to be the characteristic difference between each of the boxes we tested. They all have their charms.

They reduce noise, but I would be careful to not use the verb "filter". They are each able to pass less noise, I think mainly by imparting less of their own.

Keith remarked that the Tempus seems very sensitive to power delivery. We all bore witness to this in that it sounded great with a ~$300 Shunyata power cord and even better with a $600 Synergistic power cord.

It was the $400 Shunyata Venom V14 digital, but otherwise spot on.

I'll land this plane by saying that I, like Ken, was very happy with my 'hot rodded' etherREGEN combo but for two things. First, it only supports 100Mbit

Yes, I have been very happy with the sound quality from my ER. So much so that I'm not really in the market for another switch. I will buy a ER Gen 2 though, mostly because it will support gigabit. It can be painfully slow to copy large PGGB'd files over to the internal drives of my Antipodes server. Also, when files are in the process of being copied over, music can't be played from my NAS as there will be dropouts. The ER Gen 1 will get moved to my AV system once the Gen 2 lands.

The Tempus is clearly a stellar product and priced well relative to its performance.

Agree totally.

how long can I resist the temptation to trade up to one?

I give you about a year.

 

[kennyb123]

So what does this actually mean? Is isolation not binary?

I think he answered this by saying that "these transformers will block DC and low-source-impedance leakage. However, they DO NOT block high-source-impedance AC leakage." So there still is some leakage through these transformers. Were there not, audiophile switches might not do as much.

 

[dbastin]

It was a fun time with @jeremya and @keithc and a bunch of switches. I think we all agreed we could be happy with any of them, but there were some important differences between them.

It took us a bit of effort to land on the approach. The cable feeding the switches made very obvious difference. The Tempus lost quite a bit when fed with a Cable Matters CAT6a - even to the point where it lost ground relative to the ER fed with optical. Moving fiber over to the Tempus brought a big improvement but since the two other switches don’t have optical inputs, we wanted to avoid using optical. Fortunately Jeremy brought along a QSA Lanedri Ethernet cable. This was a big step up from optical - so much so I think we each said “wow” and laughed about it. So from that point on the QSA-L was used into the switch and the Sigma Ethernet was used out of it.

I would encourage everyone to take my observations with a grain of salt as we were pretty casual about the process. We limited the time spent listening to tracks and really only listened to a couple of them. The three contenders (PhoenixNET, Synergistic, Tempus) were each plugged initially into the wall (a spare dedicated circuit) using a Shunyata Venom V14 power cord. The ER was given an advantage initially as it was plugged into my Denali v1 using a Shunyata V14 Digital power cord. We later moved that cord over to our two favorites to see how much difference it would make. The ER used the reference clock from the REF10 SE 120 the entire time to keep things more equal in terms of retail price.

The PhoenixNET was eliminated early as it was easily bested by the Tempus with the Phoenix sounded veiled in comparison. Part of the reason also is that four switches was just one too many as far as being able to do fair comparisons. We started losing track of things. Reducing down to three helped. Also, there PhoenixNET might have been put at a disadvantage because its clock apparently needs at least 24 hours to settle back in.

The Tempus was definitely my favorite. I was sitting in the center seat when the guys switched over to the Tempus.We were listening to Diana Krall singing A Case of You from her live album. The piano sounded so much more real and beautiful that I started to tear up. I then gasped when she started singing as her voice became so clear an present.

The Synergistic switch came in second. It was still veiled relative to the Tempus, but not as badly so as the Phoenix.

Powering the Synergistic and Tempus switches off the Denali brought a nice improvement. We also had on hand the Synergistic Foundation power cord that came with their switch. This when plugged into the Denali improved the sound of both the Synergistic and Tempus switches but it improved the Tempus even more. The distance between them grew as a result.

The ER came in third. I thought it was closer to the Synergistic than the Synergistic was to the Tempus but the others might not agree. Hopefully they will post their impressions here as well.

I think the biggest surprise to me was how much difference the cable feeding the Tempus made. That Cable Matters copper cable harmed the rise times and it wasn’t subtle. I should point out that this cable had only three days of use as I purchased it at Amazon specifically for this comparison. That just reminded me of our final comparison. We tried the QSA-L directly into the K50 from the wall. It brought all its positives with it but it didn’t obviate the need for a switch to reduce network noise.

Well I think that about covers it. I think my one takeaway is that I will need to purchase a QSA-L Ethernet cable at some point. I believe that might have made as big of a difference as the switches themselves.

I am new to this thread and read most posts since the above by Kenny due his cross-reference elsewhere. There is some good discussion here. I have a few key comments to chip in about what I have found (briefly).

Synergistic Research Tranquility Pods (EM Cells) are very effective at improving SQ when used with any switch I tried, and that improvement increases with upgrading grounding, footers and power supply. I have them under ethernet devices that are not able to be changed for 'audiophile' versions and under my ER. The SR Switch and Router have these EM cells at their heart.

SR ECTs placed inside switches are quite beneficial.

Shakti On-lines on ethernet cable connectors to be quite effective, perhaps even more than the Shakti stones on switches etc

I use a Mikrotik CRS 305 as router because it is 10G and sounds a lot better than the Ubiquiti Edge Router X (1G) it replaced. There is a whole thread on Audiophile Style about Optical that covers 10G benefits.

Put caps like Audioquest on all unused RJ45 ports.

I am surprised the QSA-L etherent cable is 6m, it suggest it is 'absorbing' something unwanted enroute. I'd be curious how that compares with short cables that also absorb - perhaps the likes of Shunyata Omega.

I am not surprised to hear changing power cables to switches makes a difference.

I am not so sure veiled sounding is automatically a bad thing, sometimes that seems to be a step towards and overall improvement depending on some other tweak to 'illuminate'.

If I have time I will reply to specific comments later.

 

[KPC]

It was a fun time with @jeremya and @keithc and a bunch of switches. I think we all agreed we could be happy with any of them, but there were some important differences between them.

It took us a bit of effort to land on the approach. The cable feeding the switches made very obvious difference. The Tempus lost quite a bit when fed with a Cable Matters CAT6a - even to the point where it lost ground relative to the ER fed with optical. Moving fiber over to the Tempus brought a big improvement but since the two other switches don’t have optical inputs, we wanted to avoid using optical. Fortunately Jeremy brought along a QSA Lanedri Ethernet cable. This was a big step up from optical - so much so I think we each said “wow” and laughed about it. So from that point on the QSA-L was used into the switch and the Sigma Ethernet was used out of it.

I would encourage everyone to take my observations with a grain of salt as we were pretty casual about the process. We limited the time spent listening to tracks and really only listened to a couple of them. The three contenders (PhoenixNET, Synergistic, Tempus) were each plugged initially into the wall (a spare dedicated circuit) using a Shunyata Venom V14 power cord. The ER was given an advantage initially as it was plugged into my Denali v1 using a Shunyata V14 Digital power cord. We later moved that cord over to our two favorites to see how much difference it would make. The ER used the reference clock from the REF10 SE 120 the entire time to keep things more equal in terms of retail price.

The PhoenixNET was eliminated early as it was easily bested by the Tempus with the Phoenix sounded veiled in comparison. Part of the reason also is that four switches was just one too many as far as being able to do fair comparisons. We started losing track of things. Reducing down to three helped. Also, there PhoenixNET might have been put at a disadvantage because its clock apparently needs at least 24 hours to settle back in.

The Tempus was definitely my favorite. I was sitting in the center seat when the guys switched over to the Tempus.We were listening to Diana Krall singing A Case of You from her live album. The piano sounded so much more real and beautiful that I started to tear up. I then gasped when she started singing as her voice became so clear an present.

The Synergistic switch came in second. It was still veiled relative to the Tempus, but not as badly so as the Phoenix.

Powering the Synergistic and Tempus switches off the Denali brought a nice improvement. We also had on hand the Synergistic Foundation power cord that came with their switch. This when plugged into the Denali improved the sound of both the Synergistic and Tempus switches but it improved the Tempus even more. The distance between them grew as a result.

The ER came in third. I thought it was closer to the Synergistic than the Synergistic was to the Tempus but the others might not agree. Hopefully they will post their impressions here as well.

I think the biggest surprise to me was how much difference the cable feeding the Tempus made. That Cable Matters copper cable harmed the rise times and it wasn’t subtle. I should point out that this cable had only three days of use as I purchased it at Amazon specifically for this comparison. That just reminded me of our final comparison. We tried the QSA-L directly into the K50 from the wall. It brought all its positives with it but it didn’t obviate the need for a switch to reduce network noise.

Well I think that about covers it. I think my one takeaway is that I will need to purchase a QSA-L Ethernet cable at some point. I believe that might have made as big of a difference as the switches themselves.

Which QSA-L Ethernet cable- Gamma Revelation, Gamma Infinity, or Spectra Combo?

 

[kennyb123]

Which QSA-L Ethernet cable- Gamma Revelation, Gamma Infinity, or Spectra Combo?

Good question as I should have indicated that. I believe @jeremya mentioned it in one of his posts. It’s a Gamma Infinity. It’s the original version with a black jacket. I believe he’s getting that swapped for the current version.

 

[kennyb123]

I am not so sure veiled sounding is automatically a bad thing, sometimes that seems to be a step towards and overall improvement depending on some other tweak to 'illuminate'.

By my definition of veiled, it can never be a good thing, as my ultimate goal is for my system to get the heck out of the way of the music. If a tweak is found to illuminate then it would get substituted for something that better gets out of the way. I don’t subscribe to the “offset this by adding that” approach. I think that’s why @jeremya was able to say that my system is an “excellent lens to experience music through”.

Not everyone is after that though. I just mention it as I think it helps to explain how I stack-ranked the switches. It was essentially a ranking of how well each switch helped my system to get out of the way of the music.

 

[treitz3]

what about internals of the muon pro filter?

Hello and good afternoon to you. All you should really concern yourself with (honestly) is the end result as to what hits your ears. That is what is important.

When you hear it (after break in) is all that really matters. At the end of the day, they have proprietary items inside the box, so you can guess but will never know.

Kind of like QSA products. People clammer about knowing it's this or that but again....there are proprietary items within the product. Unless you know what they are or buy their products? You will never know.

Tom

 

[dbastin]

I have become rather sensitive to the effects of noise and I have come to realise that what is sometimes mentioned as being an improvement such as enhanced detail or bigger soundstage or more lifelike is more likely to be the symptoms of noise artefacts.

Some might call it veiled when those artefacts are removed. I have experienced reducing noise in increments so much that what initially sounded veiled, dark or dull gradually revealed more nuance and detail more clearly which might have previously been emphasised in a distorted way by noise.

I am slowly learning what un-distorted sounds like.

I also use fibre, two 25m cables, with CAT6a in parallel.

I'd be interested to know why copper in parallel to fibre is used.

the worst is a shielded cable in which the shield is grounded at both ends and acts as a noise conductor so undoes the good work of the galvanic isolation in the switch.

I know a few reputable audiophile brands that do this because its sounds best to them. I have one and compared with a competitors cable with the shield connected at only one end there is not much difference.

They reduce noise, but I would be careful to not use the verb "filter". They are each able to pass less noise, I think mainly by imparting less of their own.

I would say that each device (and cable) is both impacted by and pass on noise to a different degree, the end result at the streamer is the combined benefits of each thing's measures to reduce the noise and impacts.

 

[audiophilac1]

Hello and good afternoon to you. All you should really concern yourself with (honestly) is the end result as to what hits your ears. That is what is important.

When you hear it (after break in) is all that really matters. At the end of the day, they have proprietary items inside the box, so you can guess but will never know.

Kind of like QSA products. People clammer about knowing it's this or that but again....there are proprietary items within the product. Unless you know what they are or buy their products? You will never know.

Tom

gotcha. anyway the post (of the internals) was for fun.
As always, let their magic wow me

 

[audiophilac1]

anyone did a comparison between Ediscreation silent switch ocxo and zaiyin stargate?

i have the ediscreation on loan.. before i commit to buy - just to be sure

I'm using the ediscreation between my modem and FMC (copper cable/sablon ethernet 2020)
The FMC connects to router via fibre optic.

Don't know if the Stargate, as FMC would sound good/perform good like the Ediscreation switch? Or should I still use both Edis+Stargate (Stacked?) - but this would be quite a dream as I still need a Tempus (or something) for the front-end

 

[NigelB]

The best cable between a switch and a streamer is one which is shielded but the shield is grounded only at one end; the next best is a completely unshielded cable like Cat6; the worst is a shielded cable in which the shield is grounded at both ends and acts as a noise conductor so undoes the good work of the galvanic isolation in the switch.

I know a few reputable audiophile brands that do this because its sounds best to them. I have one and compared with a competitors cable with the shield connected at only one end there is not much difference.

Hi @dbastin , for avoidance of doubt and all that. When you quoted me earlier, you used only the (now italicised) piece after the semi-colon so may have missed the context of the very specific use case (now in bold) to which I referred which is between a switch and a streamer. This is for the reasons stated, that a network cable with the shield grounded at both ends can effectively undo the good work of the switch in reducing RFI noise. Does this same use case apply to your response? Did you try the above between a switch and streamer or elsewhere in your network?

For what it's worth, for the 10m run from router to switch I use Cat6a (shielded grounded at both ends) and then from switch to streamer I use a cable with the shield(s) grounded at only the switch end.

 

[Republicoftexas69]

So I guess you should rethink.... [Edited; please refrain from commentary abusive to other members. The WBF would like to retain members, not run them off for differing opinions.]

up to 40 Gbps for a max distance of 30 meters, while Cat 7 offers a speed of 40 Gbps for a distance of up to 10 meters. When it comes to frequency, we have a clear winner. Cat 8 cables work at a higher frequency of 2000MHz, or 2GHZ

Sure okay.

 

[NigelB]

I'm using the ediscreation between my modem and FMC (copper cable/sablon ethernet 2020)
The FMC connects to router via fibre optic.

Don't know if the Stargate, as FMC would sound good/perform good like the Ediscreation switch? Or should I still use both Edis+Stargate (Stacked?) - but this would be quite a dream as I still need a Tempus (or something) for the front-end

So you have router > optical > Ediscreation > FMC > ...

Sorry if I missed it earlier but what is the rest of your network up to the streamer please? Thanks

 

[ssfas]

Kind of like QSA products. People clammer about knowing it's this or that but again....there are proprietary items within the product. Unless you know what they are or buy their products? You will never know.

QSA 13A fuses are sold in the UK. These are safety fuses and subject to strict legal regulations and labelling requirements.

I emailed the UK distributor. He confirmed by email that they are compliant Bussmann fuses underneath that have been relabelled and subject to QSA's treatment. There is no physical modification, else they would be illegal. As QSA won't explain their magic sauce, whether it changes their performance and legality is unknown.

UK law requires the manufacturer name and BS1362 certification to be displayed. In this regard they don't comply.

As these 13A fuses are marketed for up to $10,000 each in the UK, I can only assume all other QSA fuses are prepared the same way from stock fuses. In fact this can be seen from their YouTube channel.

 

[audiophilac1]

So you have router > optical > Ediscreation > FMC > ...

Sorry if I missed it earlier but what is the rest of your network up to the streamer please? Thanks

its complicated.. i'll try.

Internet:
optical network terminal (ONT) > ediscreation switch > FMC > router

Note: I prefer fibre optic over copper transceiver. hence the use of FMC

Internal network
Router > Mikrotik SFP switch > etherRegen > Streamer (Lumin u2 mini)

Note: Fibre optic is used from from Router to etherRegen

I'm probably getting a Zayin Stargate to eliminate the ediscreation and FMC.
I intend to upgrade the etherRegen too

 

[audiobomber]

For what it's worth, for the 10m run from router to switch I use Cat6a (shielded grounded at both ends)

But that shield provides a connection between two ground points with different potentials, and will certainly result in current flowing from one to the other, something I have been advised to avoid. I believe you will hear an improvement if you add a LAN Isolator between the router and switch, to interrupt the continuity. A good one will also clean the signal from the router.