Relationship between Ethernet Switch and SQ

fredrik222

... how about an actual test and dissecting of an "audiophile" switch?

People have reported their results but you tend to dismiss them with a wave of the hand or a string of insults. Of course you could conduct your own test, couldn’t you?

But when you don’t know anything about the topic, it is easy to be caught up details that really do not matter.

Nasty insults are no substitute for reason.

@tonywinga 

Interesting video.  He shows data of how performance degrades under load. 

This is insight you just got is below foundational knowledge for the topic of Networking. 

And it doesn't really apply to switches that doesn't use a single CPU for all traffic handling under most circumstances. If you look at my Juniper EX220-C, the switch can handle 28 Gbit/s and 21 Mpps, which given the 14 ports equates to wire speed, i.e you can have every single port go 100% load and the switch can handle it. In contrast, the Switch X has a 10 Gbit/s backplane with 8 ports, so it can only handle 62.5% load on all interfaces before it dies. And these tests are with jumbo frames, not real life traffic, so it makes it most likely around 30% maximum load.

But that really won't matter if you use it for audio only. It does matter however if you use it as your primary switch and lots of other things plugged into it.

Again, this is so basic that it is ridicules, the concept of load affecting performance! But when you don't know anything about the topic, it is easy to be caught up details that really do not matter. 

@deep_333

4 hour video about two guys listening to music seems pretty useless, however, I did watch the section "some sort of conclusion" where the guy on the left says that the LHY and PURA "audiophile switches" sounds worse than the baseline, based on the timing and rhythm, so there is that.

and you realize you are the one in a cult right? You speak out against technologies and science you know nothing about, jumping up and down screaming “but I heard it, I did, I promise!”

how about an actual test and dissecting of an "audiophile" switch?

I moved the SW-8 next to the EE8.  So the SW-8 is connected to the Apple TV at the front end of the long ethernet cable vs. the back end.  I also left the computer unconnected- it doesn’t need an ethernet cable anyway.  The TV picture streaming looks the same as when the SW-8 was at the back end of the long Ethernet cable and next to the Apple TV.

The EE8 now has one short connection coming from the SW-8 and one connection to my music server.  It will be a couple of days before I listen to my stereo system.  I shut the stereo down last night so I can get some yard work done the next couple of days. 

Interesting video.  He shows data of how performance degrades under load.  I had wondered that in the back of my mind since my Audio network switch has my computer, back channel router and Apple TV plugged into it.  The EE8 switch says to use one particular port for audio because it is isolated from the rest.  I’m thinking about moving the SW-8 from the Apple TV and putting it next to the EE8.  Plug everything into the SW-8 and then have one cable from the SW-8 go to the EE8 and on to my music server.  I’ll see if that makes a difference.  (I will also try just unplugging everything but the Audio from the EE8.)

I saw an improvement in the TV picture streaming when I put the SW-8 next to the Apple TV.  My assumption was that the long, 100’ ethernet cable run picks up some noise.  But maybe it is the multiple connections into the EE8 that are adding noise.

Let us begin your rehabilitation Freddy. After you get rehabilitated, go back to your Sith lord Majidimehr on the ASR mothership and serve as the rehabilitated, repurposed, audiogon Trojan warrior.

But, watch this (all pinned timestamps, live blind tests, etc) and let your mind start polka dancing first. This is step 1 of your rehabilitation.

https://www.youtube.com/live/8HNMMksiD_Q?si=leRtfRhKfpQTjfMS

@nonoise 

No need to spend $3K for the one sold at GTT Audio but you'll have to source your own linear power supply and it won't have any of the mods in it but you'll be way ahead of the game for pennies on the dollar.

keep in mind that all communication devices are regulated by the FCC, and while they are not likely to care if you modify a switch as the worst case scenario is that you burn down your own house, it is very different with radios. It is illegal to sell and operate radios that have not been approved by the FCC, and that includes modifying existing radios. Unless of course his “modifications” to the Aruba access point doesn’t do anything , which is the most likely scenario, since any changes would need to include software update, and he doesn’t have source code access to the Aruba OS. 

@nonoise who is the dishonest broker now? 

he literally says the ground is on the backplane, and you can see that what he is running his hand over is between the SFPs and the CPU, so he is says the ground is connected to the CPU. That is what he is saying. He likely doesn’t understand what he is saying however.

Give me one article about electrical jitter? Electricity is in the analogue domain where jitter doesn’t exists. Are you perhaps confusing electrical and electronic?

it was definitely a rant. And double regulated doesn’t mean that you say it does,. It means it can handle two different ranges, for instance 200 - 240 and 100 - 120V. And he says it makes the component more stable, which it doesn’t.
 

a switch literally handles thousands of parallel threads per second, it is the only job of a switch.
 

And @nonoise, what about his ludicrous rant about isolation gadget? Pure fiction.

So, now you have proven yourself again to be a dishonest broker, for what? 

Then he says it ground canals on the circuit board, and that is complete wrong, you don't run ground from the SFP to the CPU, you run it to the ground. 

No, he doesn't. He says there's a ground plane on the back side of the circuit board and then says that the channels/ canals) allow for discrete paths for each input to the chip.

Continuing - he says "electrical" jitter. There is no such thing. Does not exist, just made up to sound good I guess?

He also mentioned "packet jitter" with "electrical jitter" and if you google electrical jitter there's tons of info on it online under that term.

Then he goes on a rant about double regulated voltage, also bunch of made up things. It does not improve stability of any component because it can handle two separate voltages. In fact, most higher end audio equipment like my Classe stack does the opposite to have more pure operation based on the voltage you use. So opposite of what he states.

He never ranted. Anyone can view the video above and see that. Also, he said it's double regulated to handle voltages ranging from 12 to 57 volts and not just two separate voltages so doesn't it perform like your higher end audio Classe which adjusts to the voltage used? 

And yes it's a less expensive design using a CPU rather than a FPGA but no one is running thousands of threads simultaneously. That would be overkill. If you google 10 bit CPU you get lots of praise for what it can do. In fact, they go on to say it's overkill. You seem to enjoy damning with faint praise a lot as the unit is just fine for audio use.

All the best,
Nonoise

@nonoise  ok, so first of, Microtik is a low end manufacturer that does everything in the CPU.

Then he says it ground canals on the circuit board, and that is complete wrong, you don't run ground from the SFP to the CPU, you run it to the ground. 

Continuing - he says "electrical" jitter. There is no such thing. Does not exist, just made up to sound good I guess?

The reason it is "clean" is because it doesn't use FPGA, everything is run to 1 CPU.  It's cheap designed, but for home use it won't matter.

Then he goes on and says that it is 10GBit chip, powerful and then says no one talks about the chip. Well, most modern designs don't use a CPU, they use FPGA, and then he talks about the throughput, yeah, every single switch worth its name is talking about it in specs, and not just jumbo frame throughput, but small frame throughput and frames per second.

he is however right about the more powerful the CPU is, the less jitter it will introduce. However, this is one of the reason why good switches don't use a CPU, they use FPGA.

Then he goes on a rant about double regulated voltage, also bunch of made up things. It does not improve stability of any component because it can handle two separate voltages. In fact, most higher end audio equipment like my Classe stack does the opposite to have more pure operation based on the voltage you use. So opposite of what he states.

"Dedicated ethernet port and dedicated isolation gadget" - oh my, where do I start here??? All made up again, according to the manufacture again, it's a LAN management port chip, nothing to do with isolation, link below.

https://www.xmultiple.com/wp-content/uploads/2021/03/XTFZ-PC13N1ZH.pdf

So, there you have it. Another "audiophile" switch debunked! 

@nonoise  no, that is not what I said. I said I know more about networking on this forum than anyone else. And so far this holds true.

Provide a link and I will explain why it is not relevant.

How am I not an honest broker? Point to one post where I wasn't able to back up my statements or claims.

You said a long time ago in a heated debate that until anyone smarter than you shows up, you're right and everyone else is wrong. That speaks volumes about your mindset. Something tells me that someone like Chris (dejitterit.com) is smarter than you and you're afraid to actually talk to the guy for fear of being shot down.

As for you claim:
In a residential setting, there is no RFI propagation through Ethernet, it is impossible. 

I googled RFI on ethernet and got pages of all manner of people discussing it and how to get rid of it. Lots of them are engineers, radio operators, etc. 

All the best,
Nonoise

And here's one on the Aruba:

Like he says, you can get the Aruba, used, over at eBay. Just make sure it's from a seller with great feedback. There's tons of them online and you can scoop one up for $15 or go for a brand new one for $125. 

Also, make sure it's IAP 225-US so it's configurable and legal to use.

No need to spend $3K for the one sold at GTT Audio but you'll have to source your own linear power supply and it won't have any of the mods in it but you'll be way ahead of the game for pennies on the dollar.

All the best,
Nonoise

@cleeds The Dunning Kruger effect clearly shows that FACTS DO NOT MATTER to people who don't know them. But again, it is ok for everyone to jump on me when I answer the OP, but when I call them out for not knowing anything about the topic, I am the one insulting?

@nonoise  How am I not an honest broker? Point to one post where I wasn't able to back up my statements or claims.

fredrik222

I am an expert, however, you fall under the Dunning Kruger effect ... I called him a BSer. Because that is the truth ... he is just another clueless guy hawking gear that doesn't do anything, at best, at worst deliberately trying to make money of people who don't know better. 

Nasty insults are no substitute for reason. They actually weaken whatever argument it is that you might be trying to make.

Here's one of Chris's videos on the Switch X:

All the best,
Nonoise 

@fredrik222  It would be a lot easier if you just called both of them and discussed it and got back to us with your findings. You've shown yourself not to be an honest broker so relying on you to educate us is a non starter.

Or better yet, let them know of this thread and how you feel about them. Maybe Bill or Chris could come on and speak their mind a bit? Others who've had been disparaged have and it helped to clear the air some. 

All the best,
Nonoise

@nonoise  so, the moderators did not like calling the guy what he is. 

I have no idea as I never read your insult (which is all you really do).

for instance, he says all electronics are contained in the SFP, which is patently false. Open ANY switch and you will see this very clearly. 

Isn't it clear that he's referring to the electronics inside the SFP that allow it to mate to the electronics in the switch? He knows there's electronics in the switch as he talks about electronics in the switch for practically the whole video. Were you drunk when you watched it?

As for everything else, I guess you didn't bother to call either of them and discuss the matter, from what you wrote.

All the best,
Nonoise

@fredrik222  You seem to have this delusion that you are an expert on this topic.  As far as we know, you are not.   You have never posted a technical article that you have published in the appropriate journal nor shown established credentials to verify your claims.  You have not even posted an audio system to show any relevant experience with audio nor have you shared your experiences, if any listening to and comparing various network switches in a competent audio system.  You just copy/paste other peoples work and then make comments with a spin on it to support your claims.  You're just a duck in a storm- "quack, quack, quack."

@nkphoto not everyone have relevant experience, and I did answer the question posted by OP, and immediately got attacked by people who have no relevant knowledge. But I guess when I am attacked I am ruining the vibe, black ain’t black, and white isn’t white, am I right?

get well soon

@tonywinga  lots people don’t like what I have to say, but I doubt they would say I don’t know what I am talking about. Same can’t be said for you, who can’t post anything to support your fanatical and fantastical claims. 

@fredrik222 Yawn.

I see you are back at the office today.  Your boss has nothing for you to do?  Really?

This is ironic:

A lot of people really suffer from the Dunning-Kruger effect in this forum, also known as cognitive bias, or the less you know about a topic the more you think you know about said topic. 

This post of yours seems to best describe you.

@nonoise  so, the moderators did not like calling the guy what he is. 

But either way, there are no measurements in that video, and the guy is clueless. He doesn't know anything about Ethernet at all as evident by the video.

for instance, he says all electronics are contained in the SFP, which is patently false. Open ANY switch and you will see this very clearly. 

And then he says "don't use wifi unless you bought it from me". Yeah, sales guy at its finest. 

Below is a link to an article describing what is in a switch, which proves that this guy is clueless. Completely clueless.

On top of that, he references a video by another guy who states why you shouldn't use wifi, and that guy "Apple TV X", correctly states that wifi has more than 10x latency, and therefore you shouldn't use it for audio. Well, it fails to account for buffering, and what the human ear can perceive. Wifi is perfectly fine for streaming audio. 

In a residential setting, there is no RFI propagation through Ethernet, it is impossible. No one has ever posted anything factual to the contrary. Lots of people have posted things they think says this, but doesn't. Dunning-Kruger....

@tonywinga 

But you know, as an engineer I always worked with data and stood up for the truth

Thank you for the laugh of the year (early still). You have never in your life on this forum posted any data supporting anything you say, even though I have asked for it over and over again.

A lot of people really suffer from the Dunning-Kruger effect in this forum, also known as cognitive bias, or the less you know about a topic the more you think you know about said topic. 

I’ve been deeply tracking audiophile switches for quite awhile, looking for lower cost alternatives to the Telegartner M12 switch which also requires special terminated cables.  I concluded:

1. Audiophile switches must be at least to the level of English Electric 8switch, and this switch is easily beaten by more expensive audiophile switches. Home improved switches like Cisco are not enough to make a noticeable sonic difference 
2. Unless you need a switch, this is only an option. The Sonic benefit can be minor- like upgrading a cable.  Often the $3-4k+ would be better spent elsewhere 
3. You need a very transparent system to hear a worthwhile difference such as very low noise floor components which are costly 
4. Audiophile switches benefit with a better linear power supply 
5. Adding more switches reportedly brings sonic benefits

So for most people, it’s better to spend their money elsewhere than on an audiophile switch.  YMMV

The rest of this, on the network side, is audio bling, unless your streamer isn’t designed well to handle any electrical noise over its LAN port. And that is only theoretical.

Chris Stephens of DeJitterIt uses older and more accurate test equipment to ferret out noise that some say isn’t in there and gets these results. Scroll down to see them at the link. Also, see this presentation:

This has been brought up before by other members who use this product and the naysayers who populate this thread never chimed in on it when this was discussed.

Granted, it’s expensive, but it addresses what some say isn’t there and shouldn’t matter. Maybe some of the naysayers can contact GTT Audio and Chris Stephens, debate/discuss this with them, and get back to the members here and report back how it went. Maybe even do a YouTube video or start a webpage about it instead of trolling on threads.

All the best,
Nonoise

you fit the mold of someone who’s close-minded.

Ad hominem attack...not a good way to make your point...this is what closes people's minds. 

Well frederik, I'm sorry you feel that way.  Perhaps if I were more of a politician...  But I'm just a retired, cranky old engineer.  I know that I would struggle selling hot potatoes to Eskimos.  But you know, as an engineer I always worked with data and stood up for the truth- even took a hit in my career once by standing up for the truth.  So you see it is hurtful when you claim I am making false claims about my own personal experiences.  Most people on these forums, like myself are genuine and want to share our discoveries and experiences with others.  Sure, we need to watch out for that wolf in sheep's clothing now and then but they are quickly found out.  

Still can't answer the questions...

@tonywinga provide a shred of evidence as I have been asking you for 8+ months per your own statements.

You keep deflecting,

Answer the questions.

@tonywinga in the last thread you defined a troll as someone who cannot backup their claims and are only out to instigate confrontation (paraphrased), and I repeatedly asked you to back up your fanatical and fantastical claims, and you couldn’t, so, again, per your own definition, you are a troll. Every time you are asked to provide any at all supporting evidence of your outlandish claims, you take to personal attacks.

So this is a somewhat odd thread but I’ll add to the pile. Noise is an issue that is hard to fathom as when it’s there, you may not realize it. When it’s reduced you certainly can.

Had used the fiber optical solution seen here in the video and it worked very well. When I added a WiiM Pro to my monoblock tube amps, didn’t think it was needed anymore.

More recently, I had switched going out from the streamer coax and thought let’s try that glass optical cable I run for video audio. It is a substantial and a lovely improvement running that glass Toslink into an RME ADI-2 DAC. The DAC can handle 192K streaming coming out of the WiiM Pro.

Had the fiber optical cable lying around and said, hey, why not give it a try. It may not be an improvement like earlier. (Also use an LPS for the second fiber converter's power and to power the WiiM Pro.) Well, it took up the performance a whole other notch and now my system is the best sounding it’s ever been. It's sooooo good. 🐸

Both of these changes offered significant improvement because noise was removed. They are both audible and with noticeable impact. Happy listening and Happy New Year!

Fredrik, you have been saying the same single message over and over on these forums for as long as 8 months- best I can tell.  Do you think you have convinced anyone that packets are packets or that the gear makes no difference?  Have you even taken the trouble to listen to some decent stereo gear and hear it for yourself?  Perhaps, you are in the group that is not sensitive to PRAT, jagged highs or care about imaging.  Nothing wrong with that; but to assume that no one else can is wrong.

There's a reason your boss probably keeps you away from people.  Your interpersonal skills need some work.  I'm not trying to be mean.  All of us technical oriented, ie. left brain people have to make an effort at improving our interpersonal skills.  Take a look at a politician.  He can convince people to use umbrellas on a sunny day as easily as you or I can discuss an engineering principle.  Don't go that hard over but learn to see that not everything in the world is black and white- in fact the world is many shades of gray.  And learn that other peoples viewpoints have value.

I've seen my designs do things I didn't think possible.  Dynamic systems are difficult to analyze and understand.  Complex systems have interactions that can cause exponential changes.  Audio draws in us technical types I think, because of the challenges and often unexpected outcomes.  That's what keeps life interesting.  

@jji666 

…

I honestly do not want to start or perpetuate an argument but analog audio logic simply doesn't apply in the digital or network domain.  It's fundamentally different.  

digital signal is carried by the analog signal inside digital cables. That is one of the reasons the sound quality varies amongst digital interconnects. Other reasons are cable material composition, noise rejection properties and connectors. Yes the bits are bits and will be transferred correctly but in audio the logic of bits are bits doesn’t always apply. Now if you’re talking about fiber optic then the analog signal concepts don’t apply. 
@fredrik222 is right with packet transfers but doesn’t understand the difference between buffering and caching which is surprising because buffering is an essential element of networking. Caching is an essential element of data processing and rendering. He fails to understand and acknowledge he doesn’t know everything and has a rancid attitude turning discussions into wars. 

@jji666 

We are here to share our experiences and simply suggesting to learn or form your opinion by experimenting. The idea is simply to try, you may or may not like result, and that's perfectly fine. This is a hobby not some scientific project where folks open to experimenting are under any obligation to provide evidence/proof for what we hear in our environment. 

You said, you are not here to troll but with a mindset of packets are packets and all else is audio bling, you fit the mold of someone who’s close-minded. 

Fredrik, you are out of your league. 

@tonywinga  no one has proven any of

your fanatical and fantastical claims, if they had people like John Swenson of UpTone wouldn’t be laughed out of every room he goes into. 

"Packets are packets and its all just 1's and 0's."

You IT guys are 40 years behind.  This discussion started when CDs first came out.  It's been proven over and over that digital audio is more than the bit stream.  If it were not the case most of this equipment today would not exist.  In spite of the naysayers, digital audio has come a long way.  I listen to my digital rig now with immense pleasure.  It competes easily now with my analog rig and exceeds in some cases.  

Until you have heard a decent audio system for yourself with and without various digital components added in, all of you IT naysayers look like cavemen claiming heavier than air craft is impossible.  

Yes, the digital domain is different.  In the analog world, less is more.  ie. keeping the signal path short and clean is the best path to great sound.  In the digital world, analog noise on the transmission lines and power supplies and jitter manifests itself as analog noise, harshness in the highs and poor imaging/sound staging.  Jitter doesn't mean much looking at text on a computer screen but it is everything when it comes to streaming.  It looks to me that the money we spend on digital audio components is mostly for clocking accuracy to reduce jitter that was not even possible in the best labs a decade or two ago.  Sure, some people don't hear or care about the difference.  One of my brothers likes his Sony boombox over my stereo system.  He's not kidding.  To each his own.  But why oh why do you naysayers feel the need to prolong a 40+ year argument?  Go buy a Teac CD player made in 1983 at a garage sale and have yourself a ball.

My latest DAC has internal reclocking on the USB in and the AES and Coax in.  It wasn't cheap but it unquestionably sounds amazing.  Another thing with this DAC that I noticed is that the noise floor got even lower.  I used to could hear the noise floor lift up as a song was starting with my old, very good DAC.  It was like a prelude to the song.   Now, the music starts from a total black background.  That took some getting used to as well- the music starting from nowhere.

Also, I run 100 feet of direct Cat 8 cable to my dedicated room. Also your choice of DAC will be critical. Reclockers and external power supplies and grounding will drop that noise floor to a black background.

I owned the Bryston SP3. I too run all my music through my internet set up (server). I use NAS storage as well as internal 8 terabyte internal drive. I run Synergistic Research Ethernet hub, their Ethernet filtering, and their Ethernet cables and power cables. It all makes a difference. The EVO Aeon music server makes the biggest difference in my opinion. This server has the closest to vinyl sound for it’s price. The Ethernet hub is one of the best you can buy but many audiophiles could not justify the price. It is certainly very musical and much more versatile than my previous reference turntable but vinyl still sounds better to me.