Audiophile Network Switches for Streaming ... really ?

[TheFlash]

So your suggestion is that shorter ethernet cables 'sound' better?

Basically yes.
My suggestion is that a multi-metre length of network cable can pick up a substantial amount of RFI/EMI noise so the more of this you can put on the router side of your switch the better. So if you’re replacing an 8m straight run of one cable with a switch plus two cables, you will maximise the sonic goodness if you go router-8m-switch-1m-streamer vs going router-1m-switch-8m-streamer.

 

[TheFlash]

An interesting experiment is to configure your system so that it can play without a network and then see if adding the network reduces sound quality.

Absolutely this.

All a switch is (or should be!) trying to do is to get as close in noise terms to “no network” as possible- assuming your kit has some sort of buffer so the music doesn’t stop as well!

 

[Funk]

COMPLETELY misses the point. Everything you say about how the data gets to your router is right. That’s all entirely in the digital domain, and bits is bits.

Try a £20 switch just before your streamer. It might fool your unreliable ears! Seriously, this is what I tried earlier this year and then tried to work out how and why. It’s a classic hypothesis-and-test approach.

Your ears are lying to you, simple as.

No switch (£20, £200 or £2000) can alter the sound as it passes the digital data stream to the playback device. The data either arrives correctly or it doesn't - the switch has no idea that what's transmitting happens to be audio, let alone be able to influence it positively or negatively in the analogue domain.

Human ears are not accurate measuring devices and the brain is easily tricked.

 

[Fourlegs]

The argument that 1’s are 1’s and 0’s are still the same 0’s no matter which switch is used is something which is dragged out of the cupboard on a regular basis in these discussions and yet those of us who have tried an ‘audiophile’ switch in our systems would be very careful about dismissing their possible advantages. I tried an Innuos PhoenixNET switch in my system and having heard it there was no way I was going to return it to the dealer. This was not just with music streamed from the internet via Qobuz but was also the case with locally stored ripped files on my streamer / server (an Antipodes K50). In my case I already had an EtherRegen in the system and I was comparing that to the PhoenixNET and also compared it to having the K50 directly connected to my local network and the internet via a standard Cisco switch.

What I did was to set a music file playing from the hard drive on my streamer/server and then pull the ethernet cable from the back of the K50 mid track to see if I could hear any difference. With the PhoenixNET connected I did not hear any difference when I pulled the ethernet cable but with the other two switches the sound changed when the ethernet cable was pulled. For me that was a quite compelling demonstration of the effect of the PhoenixNET.

On the basis that the PhoenixNET effect might be due to it removing overlaid noise in the ethernet cable I also tried an alternative of inserting a length of fibre from the Cisco switch and then using a fibre to copper cable ethernet converter (Piccy below). This seemed to work well with the caveat that although the unit itself is cheap its effectiveness seemed very dependant on the power supply so the overall cost of the whole package of ethernet media convertor + power supply soon became not cheap. Using the supplied smps with the media converter the effect was not all that impressive.

I have been asked many times on FaceBook user groups and similar groups about the Innuos PhoenixNET and each time all I can suggest is that anyone interested should try it in their own system and make up their own mind. So far everyone I know who has tried one has decided to buy it and that is also the decision I came to because I think the price justifies the improvement I can hear.

The PhoenixNET was taken to a friend’s house and I think that it possibly made more of a difference in his system than in mine. Also in my friend’s system the PhoenixNET was compared to another audiophile switch he uses which is designed around minimising noise transfer and that switch was doing a creditable job in his system whilst costing quite a bit less than the Innuos. So although I am concentrating on the PhoenixNET purely because that is what I have it is clear that there are other products which also seem to work to varying degrees and at varying price points. It is therefore a matter of balancing the effectiveness of the device in ones own system judged against the cost but what is not in doubt as far as I am concerned is that these things are capable of sounding different.

For the avoidance of doubt, I have ‘no dog in the race’ with audiophile switches and do not sell any. However I readily admit that my mind is already susceptible to the suggestion that noise overlaid on digital signals can affect sound quality because that is the basis of the WAVE digital cables which I manufacture and sell. I bought my PhoenixNET as a retail customer from my local Innuos dealer (@Musicraft of this parish as it happens and I am happy to recommend Rick).

The fibre to RJ45 converter that I have also tried connected to a Cisco switch via fibre.

 

[Funk]

Basically yes.
My suggestion is that a multi-metre length of network cable can pick up a substantial amount of RFI/EMI noise so the more of this you can put on the router side of your switch the better. So if you’re replacing an 8m straight run of one cable with a switch plus two cables, you will maximise the sonic goodness if you go router-8m-switch-1m-streamer vs going router-1m-switch-8m-streamer.

This is so far wrong it's ridiculous.


Anyway, you guys spend your money on pointless networking gear if it makes you happy. It's just sad seeing people getting conned, that's all.

 

[TheFlash]

Your ears are lying to you, simple as.

No switch (£20, £200 or £2000) can alter the sound as it passes the digital data strean to the playback device. The data either arrives correctly or it doesn't - the switch has no idea that what's transmitting happens to be audio, let alone be able to influence it positively or negatively in the analogue domain.

Human ears are not accurate measuring devices and the brain is easily tricked.

No, no, no. You’re talking about the digital domain and if you limit the possibilities to just that then you’re absolutely correct!
But that’s not the domain in which a switch can do much at all: the digital data arrives in good order and leaves in good order. I have never asserted otherwise.
Please consider borrowing a £20 switch if you can’t bring yourself to buy one, and install it just before your streamer. It will save an awful lot of correspondence.

 

[TheFlash]

This is so far wrong it's ridiculous.


Anyway, you guys spend your money on pointless networking gear if it makes you happy. It's just sad seeing people getting conned, that's all.

No-one is conned, not me anyway. I have worked in the digital domain - IT industry- for decades and think I understand how digital data moves around. The impact of a network switch is NOT in the digital domain. Seriously, if you have a long-ish network cable to your streamer today, borrow a £20 switch to stick just before your streamer. Then we can talk about the mechanisms by which it can possibly do what you’ll probably hear!

 

[Funk]

I've solved every ethernet cable-related issue by going wireless. No cables to be corrupted by that nasty 'interference' stuff..!

I. R. Jenius.

 

[Dom500]

A Cisco network switch between my sky router and allo streamer made a difference that was plain to my ears. I guess I could have been imagining it but at £28 I wasn't hoping for much and I've sent back stuff in the past for making no discernible improvement. I can't afford to spend loads of hifi but honestly if it had done nothing positive I would have confined it to the cupboard, it's definitely not there for looks or pride of ownership.
Rest of the system is an RME DAC into a Leak S20 and NS1000.

 

[Fourlegs]

I've solved every ethernet cable-related issue by going wireless. No cables to be corrupted by that nasty 'interference' stuff..!

I. R. Jenius.

indeed U. R. Jenius.

Wot a brilliant idea to activate a wifi radio transmitter inside your hifi device.

As it happens both of my streamers (Innuos and Antipodes) are designed by the respective manufacturers as wired ethernet only just for that very reason.

 

[mansr]

Anyone claiming that a network switch can alter the analogue result of its (dumb) digital throughput has no idea how a network works. As for 'reducing noise' to the port of the audio device....well all I'll say is that if your playback device is so susceptible to any such 'noise' that its sonic output is affected then it's fundamentally shit, badly designed and should be put in the nearest bin.

It's a stupid notion for (at least) two reasons. Firstly, all Ethernet interfaces are transformer coupled. Whatever nasties the switch blocks would also be blocked by the transformer on the device itself. Likewise, whatever gets through the isolation on the device will also pass through the switch. Secondly, any noise getting in through the Ethernet interface is most likely dwarfed by noise generated within the audio device. If the analogue side of the device can't cope with the former, then it will have a much bigger problem with the latter. Just bin it.

 

[Funk]

indeed U. R. Jenius.

Wot a brilliant idea to activate a wifi radio transmitter inside your hifi device.

As it happens both of my streamers (Innuos and Antipodes) are designed by the respective manufacturers as wired ethernet only just for that very reason.

It's OK, the antenna's on the outside of the box.

 

[paulfromcamden]

One thing that strikes me with all of this (and I guess also to some degree with the Cable Wars) is that digital audio can be so good now for so little money that folk who in yesteryear would have been tweaking every last ounce of performance out of their LP12 have limited options to fettle their source.

Digital folk can't spend their afternoons worrying about tracking force and cartridge alignment so we get a renewed emphasis on stuff like cables, network equipment, clean power etc.

 

[kernow]

This thread made me want to go lie down

 

[garyi]

Always impressive that improvements are wrought from more and more expensive gear. But in any audio device we are all allowed that one bargain to prove it’s not just expensive gear!

It would be funny if people were not getting relieved of a lot of money.

It’s clear the one Linus looked at is a straight up scam, I am confident anyone who watched that would concede.
As for other manufacturers well at least they make a nice case for it I suppose.

 

[hifi-dog]

How long are cables from router to EE8 and from EE8 to streamer please? And what is attached to EE8 besides router and streamer?

router to ee8 2m, ee8 to streaming bridge, 5m. ee8 also has innuos zen mini as roon core and nas attached.. The ee8 is with the tv on the other side of the room as no more plugs at hifi end.

 

[Funk]

Always impressive that improvements are wrought from more and more expensive gear. But in any audio device we are all allowed that one bargain to prove it’s not just expensive gear!

It would be funny if people were not getting relieved of a lot of money.

It’s clear the one Linus looked at is a straight up scam, I am confident anyone who watched that would concede.
As for other manufacturers well at least they make a nice case for it I suppose.

Reminds me of the Lexicon scam where they literally just dropped a $600 Oppo player (not THX Certified, nor capable of being so) into a machined case and slapped a $3,500 price tag on it. Somehow it managed to magically get THX certification somewhere along the line too. A reviewer waxed lyrical about how much better the Lexicon was, even once he was told it was the same innards as the Oppo. 'Dr.' Something-or-other. Anyway, credibility: gone.

I can't think of any other industry that's so rife with lies and bullshit for kit that's demonstrably no better than the cheaper options. Or that people believe it. At least with things like obscenely expensive watches or hypercars you can see where the engineering goes.

One thing that strikes me with all of this (and I guess also to some degree with the Cable Wars) is that digital audio can be so good now for so little money that folk who in yesteryear would have been tweaking every last ounce of performance out of their LP12 have limited options to fettle their source.

Digital folk can't spend their afternoons worrying about tracking force and cartridge alignment so we get a renewed emphasis on stuff like cables, network equipment, clean power etc.

Spot on really - and the companies that perpetuate the lies are just petrified their cash-cows are dying. In many respects they're fortunate that the many still believe the lies, although that will change with time. Audiophiles are a dying breed.

 

[Fourlegs]

One thing that strikes me with all of this (and I guess also to some degree with the Cable Wars) is that digital audio can be so good now for so little money that folk who in yesteryear would have been tweaking every last ounce of performance out of their LP12 have limited options to fettle their source.

Digital folk can't spend their afternoons worrying about tracking force and cartridge alignment so we get a renewed emphasis on stuff like cables, network equipment, clean power etc.

Of course there is an element of truth in what you say.

On the other hand if digital devices such as servers and streamers are ‘only’ passing through digital information then they would all sound the same when played through the same DAC and yet that is demonstrably not the case. Hell, even different bit perfect playback software and firmware on the same device can (and do) sound different.

 

[TheFlash]

router to ee8 2m, ee8 to streaming bridge, 5m. ee8 also has innuos zen mini as roon core and nas attached.. The ee8 is with the tv on the other side of the room as no more plugs at hifi end.

Thanks. It would be very interesting indeed to experiment with router to ee8 5m then ee8 to streaming bridge 2m (or less if you can make that work)… though it sounds like you’d need to invest in one of those 2 or 3 way mini-bricks to get an extra power socket! If you can’t plug the Zen Mini etc direct into the router, I’d personally out a cheap switch at that end and move EE8 to hifi end which is where it was designed to go (hence included 1m cable).

If you can ever be bothered of course!

 

[TheFlash]

It's a stupid notion for (at least) two reasons. Firstly, all Ethernet interfaces are transformer coupled. Whatever nasties the switch blocks would also be blocked by the transformer on the device itself. Likewise, whatever gets through the isolation on the device will also pass through the switch. Secondly, any noise getting in through the Ethernet interface is most likely dwarfed by noise generated within the audio device. If the analogue side of the device can't cope with the former, then it will have a much bigger problem with the latter. Just bin it.

It might look stupid to you (I’m not sure I’d be quite so insulting myself) and I can see why, but this is why I think it makes sense to test the theory of “this can’t possibly do anything so why should I even try?” with a £20 switch installed in the right place. Sometimes you’ve got to put down your copy of What Car and test drive the thing. In fact, I think £20 might be an exaggeration and you can pick up a Zyxel GS108B for £17 off Amazon.

I was hugely sceptical just 9 months ago. No I won’t be trying cable lifters or grounding boxes using the same “just try it, you never know” logic; I might be stupid but I’m not daft!