Thread: Ethernet switch box

Originally Posted by TheOctopus

What is even funnier is a few people getting together and pulling a fast one on international wine competition. People attending the show and judges themselves voted their water, vinegar and juice concoction with a fancy bottle label of a new company into 2nd place after a prestigious wine maker. It's human nature, intriguing labels and group confirmation bias really works.

There is indeed a lot of subjective (i.e., scientifically unsubstantiated) claims in this hobby. Buyer beware.

Originally Posted by Randy Myers

I actually have a degree in Psychology, which is really weird for a computer programmer data guy .... I really thought that Freud was a lunatic though .

I thought about what Freud would say about this hobby. It's really simple.

From the internet:

If we wish to understand our child, we need to understand his play. Freud regarded play as the means by which the child accomplishes his first great cultural and psychological achievements; through play he expresses himself. Freud also noted how much and how well children express their thoughts and feelings through play. These are sometimes feelings that the child himself would remain ignorant of, or overwhelmed by, if he did not deal with them by acting them out in play fantasy.

Child psychoanalysts have enlarged on Freud's insights, which recognized the manifold problems and emotions children express through play; they also have shown how children use play to work through and master quite complex psychological difficulties of the past and present. So valuable is play in this connection that play therapy has become the main avenue for helping young children with their emotional difficulties. Freud said that the dream is the "royal road" to the unconscious, and this is true for adults and children alike. But play is also a "royal road" to the child's conscious and unconscious inner world; if we want to understand his inner world and help him with it, we must learn to walk this road.

The road of an audiophile is not an easy one...


Us guys are nothing more than grown kids who still love to play with their toys.

Originally Posted by Puma Cat

A Melco S100 is a good switch, but...it's 2200 bucks.

The Melco is expensive but sounds a lot better than the eR. eR might have all the gizmos and tech which is good on paper and marketing but Melco beats it in sound. I have one and know it. Mod the Melco with a PF clock and you are into a different territory. The proof is actually in the listening.

Originally Posted by Puma Cat

An UpTone Audio EtherREGEN is a much better value at almost 1/4 the price: $640. And you don't need to pop for an external linear power supply for it, either.

You are mistaken - the eR just sounds better with a better power supply. Same with Melco. All the high-speed isolator chip used in eR to separate out the data/power/ground between A and B is really good on theory but is different when in practice.

Originally Posted by Puma Cat

Don't get me started with ASR. They don't even do MSAs on their measurement system.

In fact, my guess is they wouldn't even know an MSA if it smacked them in the face.

Anyone doing meaurements without qualifying the measurement system with a statistically valid MSA cannot make any meaningful statements nor draw any meaningful or statistically valid interences about their measurements.

Has ASR ever done MS repeatibility validation? No.
Has ASR ever done MS reproducbility validation? No.
Has ASR ever done measurement system*operator variance analysis? No.
Has ASR ever done measurement system*part variance analysis? No.
Has ASR ever demonstrated smallest statistically effective measurement increment? No.

Gimme a break...I used to teach this stuff professionally...

Sorry, ASR has virtually no credibility from me. Those guys were biased about EtherREGEN and trashing it BEFORE it even shipped. Sorry, but that is NOT science. Real scientists don't bias themselves, and in particular, they don't bias themselves in the complete absence of data. I was scientist my entire professional career, and these guys give science a bad name.

The best way to deal with ASR is to provide your (or manufacturer's) own measurements and prove them wrong, wherever measurements are meaningful. It should be that simple, instead of constantly arguing. If you look at ASR, there are a some DAC manufacturers who have provided measured data for their products which matches with ASR, Matrix Audio is one such....

Originally Posted by Kingsrule

The forthcoming Taiko Audio switch will probably be a winner. Ground up design all for sonics.

I wouldn't spend a nickel for some modified switch....

Taiko switch is really interesting when they announced over at WBF. I would opt for one if they are reasonably priced. The JCAT M12 and PF Melco indeed has raised the bar significantly....

Originally Posted by Puma Cat

When you say, "no improvement", what measurement did HiFi Review do that showed "no improvement"? Do you have a link to the review, by any chance?

I'm curious if they measured the Melco for its ability to block high-source impedance leakage current and its impact on threshold jitter.

Pumacat, I hope you understand what you are saying....You can't have the cake and eat it too There is NO measurements that shows eR "ability to block high-source impedance leakage current and its impact on threshold jitter." whatever that means.

Originally Posted by Puma Cat

In fact, I'm doubtful they even have the metrology to measure high-source impedance leakage current and its impact on threshold jitter. AFAIK, the only person in the world that has this metrology is John Swenson. That's because he designed and built himself just to be able to do this analysis.

You are mistaken again - JS has been building the test equipment for almost 3yrs now. Forget the measurements, there is no ETA when it will be complete or if ever it will. The eR design is based on conjecture and mere theory. I don't mean that the theory is incorrect but it lacks evidence so far. All we know today is injecting the eR into the network chain really uplifts sound quality and I can vouch for that. I use one in my chain as well but I use other switches in the chain as well.

Neither do I have my audio traverse the network. All my files are sourced locally but the server is still connected to the network and sends/receives packet (for example Roon UI), however minimal they are. The good quality network switch still affects the sound quality but not as much if the audio had traversed. A good way to test is load an album or some tracks and listen with and without the network cable connected. The sound when the network cable is disconnected is the best your server can do.

Originally Posted by mep

Are you guys that are using switches doing it because you need more Ethernet ports?

nope, only for sound quality. For Audio, you need at most 4 ports to work with - server, streamer/endpoint, NAS and upstream. I only need two - server and upstream. So Ethernet ports are not an issue as all them has atleast that many ports to start with.

Originally Posted by Randy Myers

Hmmmm... Never tried that... I guess I simply assumed that if audio files were not travelling through the network then the occasional Roon traffic would not be an issue.

On my machine this might be different as well since the motherboard has a wifi chip built in and utilizes an external wifi antenna for Roon's network connection. There is no hard wire connection to the system what so ever. Even the Oppo has it's own external wifi adapter for the once in a blue moon that I do a SACD rip.

try wired Ethernet but it takes some attention to get it right. wifi controllers are generally noisy, unless specific attention is given to isolate it like Auralic does.

Originally Posted by Randy Myers

Now if you are considering an audiophile level router... that is a different discussion.

can you elaborate on this?

Originally Posted by mep

I am assuming you have a router that has all the ports you need and you are adding a switch to the network because some people think that adding an unnecessary piece of gear in the network will improve the sound.

even if you have sufficient ports on the router, many have found

router -> audio server < router --> audiograde switch --> audio server

these are not objective but subjective opinions, off-coarse. In both the cases, your server gets all the packets/bits intact in the same way but the two setups sounds different (and there are lot of reasons and hypothesis behind it). This is why there are proliferation of audiophile switches in the marker today

Let’s say that you get your music from a music service like Qobuz, Tidal, etc.

The signal (audio) gets to the router after traveling thousands of miles from wherever. The router is just a generic appliance provided by your internet service provider.

Then you connect your router to this switch device which connects to your audio equipment. Is this switch device supposed to “fix” whatever you got from the internet? Is there any technical proof that this is the case? (I am talking objective proof not some “I hear” stuff).

Originally Posted by nicoff

Let’s say that you get your music from a music service like Qobuz, Tidal, etc.

The signal (audio) gets to the router after traveling thousands of miles from wherever. The router is just a generic appliance provided by your internet service provider.

Then you connect your router to this switch device which connects to your audio equipment. Is this switch device supposed to “fix” whatever you got from the internet? Is there any technical proof that this is the case? (I am talking objective proof not some “I hear” stuff).

Those bits and electrons are very picky. The bits arriving through the internet will either be very happy when they arrive and welcomed by an audiophile grade router or very turned off and make your music sour.... Same with electrons. Electrons travel for miles and miles and when they arrive, you better have a fat and expensive power cord ready for them or they will tell the other "arriving electrons" as they are leaving back to the power station and then your music will be in real trouble!

Originally Posted by Randy Myers

Actually many routers are after market and not provided by the ISP. The only thing that my ISP provides is the modem. I personally use an Asus higher end model router.



Exactly

Maybe it is semantics. I believe that the “modem” provided by the ISP is also a routers. Your switch(es) connect directly to the “modem” provided by your ISP.
So my question still remains unanswered.

Originally Posted by Jack

The ISP's call the combination unit's "gateways" and they are routers and modem's in one box sharing one SMPS. With both my ISP's you can go either way but you will get better performance with separates with their modem and your own router.

Thank you! That is what I thought: that the devices provided by the ISP are a modem and a switch.

So my question is still: how can can you better performance using “their modem” and your “own router” (or switch) ?

Again looking for technical (i.e., non subjective) proof. Thanks!

Originally Posted by Devg

T
You are mistaken - the eR just sounds better with a better power supply. Same with Melco. All the high-speed isolator chip used in eR to separate out the data/power/ground between A and B is really good on theory but is different when in practice.

You're inaccurately mis-characterizing what I said. I DID NOT say that the ER did not sound better when using a good LPS, I said that you don't need one to use it. The two are completely different statements with different meanings.

I agree the ER sounds better with a good LPS; I use an LPS-1.2, but that does not mean using an external LPS is necessary to obtain nominal performance from an ER.

Guys, my question is still: how can can you better performance using your ISP provider modem/router and your own router and/or switch?
I am looking for technical data not subjective (“I can hear it”) stuff.
Thanks.

Originally Posted by Devg

Pumacat, I hope you understand what you are saying....You can't have the cake and eat it too There is NO measurements that shows eR "ability to block high-source impedance leakage current and its impact on threshold jitter." whatever that means.

There are no PUBLISHED measurements but that does not mean that John Swenson does not have them. These are completely different things.

As for threshold jitter, you can read what that is here: https://tinyurl.com/yylwak9b

Originally Posted by Devg

You are mistaken again - JS has been building the test equipment for almost 3yrs now. Forget the measurements, there is no ETA when it will be complete or if ever it will. The eR design is based on conjecture and mere theory. I don't mean that the theory is incorrect but it lacks evidence so far. All we know today is injecting the eR into the network chain really uplifts sound quality and I can vouch for that. I use one in my chain as well but I use other switches in the chain as well.

No, I'm not mistaken again, you are inaccurately mis-characterizing what I said again. I didn't say John wasn't building the test equipment for 3 years now; he started building the bespoke test equipment in or around October 2017, when he discovered the existence of high-source impedance leakage current. The entire devleopment story is published at Audiophile Style in detail; you can read it there.

As for pictures of the measurements, he may publish them...or not. He's very busy with new projects and product under development, and he's virtually under no obligation to publish photos or graphics of the measurements depicting high-source impedance leakage current and its impact on edges for a product that has been in the market for over year. The advantage of the ER's design and functionality is not based on conjecture and theory, its a real design embodiment that is fully realized as functionality, but John is in no way obligated to publish his metrology, his measurement system, "evidence" or his core competencies that underpin the design embodiment. He could treat them as a trade secret if he wanted to. And he would be fully within his rights to do so.

Do you think that dCS publishes all their engineering developments and the underpinning of their core competencies?

Guess what? They don't.

Originally Posted by mep

You are never going to get a straight answer to your question because no one here on this forum can "prove" that somehow adding a switch which wasn't necessary in the first place can somehow make your digital rig sound better. In order to do that, you would need to have some serious test gear and know how to use it. Then you would have to prove that your measurements that implied something needed fixing through an audiophile switch are valid.

In summary, we are reduced to people's feelings of what they heard when they introduced digital dingleberries into their digital rig.

You'll get a straight answer from John Swenson.

Reduction of jitter (including threshold jitter) and timing errors from phase noise as well as ground plane noise is audible. The benefits of adding a better swtich are just as audible as adding a better clock, and I don't see folks stating that adding a better clock brings doesn't bring an audible improvement. I'm sure some folks here have those nice Cybershaft and Esoteric clocks.

John Swenson has the test equipment. He built it, so he knows how to use it. He used to design Ethernet chips, PHYs, flip-flops, switch components, power supplies for the better part of 40 years at LSI, Inc, Broadcom, Cisco. It's accurate to say he know more about this any of us.

Originally Posted by W9TR

I would like to see a measurement methodology developed to determine if these devices actually work.

Let me see if I have the technical argument correct. High oscillator phase noise in consumer internet switches impresses a ground current/voltage in the DAC’s digital input circuitry. This ground voltage then changes the point at which the digital input circuit in the DAC determines if the incoming data is a 1 or 0. This maninfests itself as a timing error aka jitter, which in turn shows up in the DAC output as temporal ‘spreading’ that creates distortion sidebands in the DAC’s analog output.

Noise from smps and lps impressed on the DAC ground would also have the same effect.

So this should be measurable by looking at the jitter spectra of the DAC’s analog output.

No voodoo secret methodology needed. Just good measurement system design.

The big variable here is how the DAC handles this ground noise induced jitter.

You have it mostly correct, Tom, but not entirely.

Read this white paper: https://cdn.shopify.com/s/files/1/06...f?v=1583429386

Originally Posted by rbbert

Is this fundamentally different from attention to AC power? There we use better outlets, power conditioners or regenerators, different cables, etc, mostly on the basis of subjective impressions or objective measurements with limited or no data on their relationship to the sound.

Well, yes, partly. It is partly different from attention to AC power, but certain parts are similar (e.g. ground plane noise). Power cords don't impact timing the nanosecond domain (which are quite audible). It's fairly complicated, read John Swenson's white paper linked here (twice).

Originally Posted by rbbert

I have no personal experience since my streaming is currently limited to free Spotify on my phone, and that only to sample new music. I do use "generic" Ethernet to my network for control purposes and notice absolutely no difference whether or not it is connected.

You don't notice a difference because all consumer Ethernet switches are crap for digital streaming audio applications.

At a fundamental transfer function level, it's no different than using generic black power cords(and the garbage they bring) for powering components. This "discussion" is virtually no different than the one folks were having 15 years ago that power cords don't matter. And, we know they do.

It's real easy: order an EtherREGEN and put it in. If you don't think it brings an audible improvement that is worth $640, for ~eight bucks you can return it via USPS Priority mail for a full, no-questions asked refund. 99.97% of UpTone's customers (over 2000 have been sold to date) have found it brings an improvement worth the $640 price.

If power cords make such a big difference other than being of proper gauge and well shielded (they don't carry audio signals), should a specific power cord be used in all audio review magazine measurements? How or why would anyone trust a "measurement" of an amplifier, a DAC or any other component that is routinely measured as part of a review if a stock power cord was used. What if a power cord that is not in my system was used? Can I still trust the measurements? Asking for a friend...

Originally Posted by CPP

And then you have the identifier that always seems to come back, which power cord, which cable this or that and did you try the one that cost 3 or 4 times more and/or how do you know the measurements can be trusted and and on what equipment was it measured on etc.. etc.. etc.. a never ending circle of doubt when it comes to all things audio. .

The precision test instruments capable of measuring to -120dB of noise floor. Should they have upgraded power cords too? Because we could be living in a universe where everything is completely off and we will never reach Mars if we are measuring everything with stock power cords... Oh the horror!
Are Elon Musk and NASA even aware of the untapped potential here?

Originally Posted by Puma Cat

.....

It's real easy: order an EtherREGEN and put it in. If you don't think it brings an audible improvement that is worth $640, for ~eight bucks you can return it via USPS Priority mail for a full, no-questions asked refund. 99.97% of UpTone's customers (over 2000 have been sold to date) have found it brings an improvement worth the $640 price.

Doing some research about this product, I found this:

UpTone Audio EtherREGEN Switch Review | Audio Science Review (ASR) Forum

I think I will save my money...

Originally Posted by CPP

OH yes, all bases have to be covered for a measurement to be of worth and then is the power into the building even verified, and then what about the power companies, the building grounds, and has all the equipment been synced to the World Clock, Oh the horror is right !

Me, man I just listen to the music. Thank God all of those years at Bells Labs kept me sane.

Same here. I just enjoy music. The question of what power cord or interconnect I should get next is simply a non issue for me these days. I am not using a lamp cord or a zip cord by any means but chasing the latest and the greatest just doesn't turn me on like it used to... Having an electronics background myself, it is also enough to know what a power supply does as well as what shielding does to a power cord and how it all ties in with UL listing and fire safety... etc... Let's just say I'd rather hear what John Coltrane was trying to say through his sax than what the latest and greatest power cord of the month is.