I would have suspected #3 to be the best. Optical can have issues with short runs so better to keep the optical cable longer.
Yes, very true. And perhaps a harbinger for the downfall of mankind. One can only pray that science, and additionally, the capacity to perceive and understand at a level that transcends the tricky dualistic mind, will provide an antidote.
Well said. The scientists should always win the argument with proven facts but sadly the politicians and bullies will never let it happen.
1 Like
Lately I’ve been reading about the EtherRegen and contemplating ways it might help my system. There are tons of positive reviews all around. To my understanding of the Uptone website, the ER is designed to substantially reduce noise and jitter along the Ethernet chain.
This strikes me as a very focused task. It’s not as complicated as an amp or DAC, it’s just a tiny box that does one thing, substantially lower noise and jitter. This ought not be a matter of subjective listening, but rather a basic, easily provable fact. Noise is fed into the ER, and what comes out is much less noisy.
I subscribe to Hans B, the Dutch reviewer on YouTube. He gives high praise to the ER. He has come under fire from some network audio engineers claiming Hans has no idea what he’s talking about when it comes to switches. Some chap has gone so far to post a YouTube video of tests showing the ER does nothing to lower noise, and another video counteracting, point by point, Hans’ claims.
I understand the ASR debates, and it’s not my intention to go there. But we have a very knowledgeable group on these forums. Has anyone here with the engineering skills and abilities, with no axe to grind, actually tested the ER to see if it does indeed substantially lower noise?
Whether or not the ER improves SQ is a separate issue. I’m curious solely on the question of noise.
Has UpTone Audio posted any data showing how the EtherREGEN reduces noise? Until they do that, and they promised they would, it is just snake oil…
2 Likes
I have lots of experience with Mikrotik routers. Each and every aspect of the router is customizable, including the “role” of each port (LAN, WAN, linked in software bridge, isolated, linked in hardware switch chip). You can do load balancing, create VLANs, put on create as many networks you want, define routes, put almost any firewall rule etc). So your SFP cage may act as wan link, as LAN port integrated in bridge, as LAN port isolated from others and so it goes.
Everything is done through their proprietary OS. Very, very steep leaning curve. Including programming through a terminal of their won, alongside a quite rudimentary (and not at all intuitive) GUI. As far away from consumer-grade router as one cam be.
but once you get accustomed to it, you won’t take anything else, for the flexibility and option it gives you. But be ready for a rough software ride.
1 Like
That actually sounds pretty good 
Had a draytek 2925 for several years, just works and allows VLANs and random firewall stuff, it is however somewhat quirky, CLI and GUI.
Maybe I shall give mikrotek a look when it comes time
Shame it does t have two cages but I expect they have one that does - my managed switches all have dual sfp option, be nice too install a fibre ring in the house.
They have all types of routers.
One of the features I like most is to configure all hardwired AP through the main router (they call it CAPsMAN). You put the same SSiD, security profile etc, then you configure for AP to drop the connection once the signal gets below a certain threshold (say -80db). Then you get a “hardwired mesh” without the inconvenience of the usual mesh setup. If you spend the time it takes, it rocks.
Have a look at their wiki, just to see how powerful and flexible it is (that’s all the documentation you will get).
Sounds good to me, once my current project is done I shall update the network, it’s been due a while.
My son will not be pleased - he thinks he runs it now, 
(He actually does but since I built it, I get first dibs!)
You would have suspected that would be best eh? That is what I had until I heard “Get rid of the switch!” . But YOU knew it was best. Sigh.
1 Like
When did I say get rid of the switch to you? I said move it away from the listening room.
Sigh.
Out of the options presented #3 was the one I thought would sound best, I didn’t say I thought there would be other options that might sound better or not.
Sigh v2.0
+1.
Streamers use the stateful protocol TCP. Sure you can get packet loss, that’s why they buffer (like the portable CD player I has as a kid).
Extra network devices in the path can only introduce latency and point of failure.
Fear not. I put a filter in place to solve the problem.
I just find it unbelievable that a company making a product who’s only purpose is to lower noise, would not provide solid, 3rd party verified proof that it does that simple task. From what I understand, testing such a thing is not difficult.
They’ve sold over $1M worth of EtherRegens, all without testing? Huh? It may all be innocent. I can’t say. But with $1M in sales, there could be at least an unconscious reason to get to testing, er, manana.
Alex is a very very nice person, I can’t believe it’s a con. But the lack of test results, even their own, is peculiar.
And what’s even more unbelievable is that so many on this forum, and on AS have bought the EtherRegen not seeming to care one lick if it reduces noise. That whole thought process is baffling to me.
I’m not knocking the ER, it may be great, but testing does matter, which is why Stereophile does them.
My unscientifically proven theory on devices like the ER, from other companies as well, is that to add the device you are also adding likely 2 sets of cables and a power supply, plus the device itself. Even if the device does scientifically nothing, adding 4 things to the chain is likely to alter the sound in some fashion. And it’s that alteration that audiophile addicts are hungry for in the never end ending quest for non-existent perfection.
If only we had a world class digital audio expert to weigh in, and perhaps volunteer to test the EtherRegen? If only the PS Audio forums had such a character, who is so free with he wisdom and advice 
(To clarify, I’m not in the “it must test good to sound good camp”. Amps, DACs and Speakers are far more complex. But a device that only does one thing, reduce noise, ought to be held to a rigorous testing standard)
I think that’s part of the hobby - new shiny toy that you can get excited about and then convince yourself it was worth the cash.
After a few beers I looked at it, then I sobered up and had a chuckle.
1 Like
Well, to be fair, the kind of noise the EtherREGEN is suppose to eliminate IS difficult to measure. It is supposed to eliminate clock phase noise and John Swenson was supposed to be building a device that could measure that noise so they could verify the device did what they hoped it would do.
I don’t visit the Computer Audiophile. or whatever it is called now, forum anymore so I don’t know what their excuses are now. But they really should come up with data to backup the claims.
1 Like
I don’t get how clock noise on the network could impact the TCP protocol from an audio feed perspective.
As long as the data gets to the streamer and the onboard NIC can feed the buffer, all the audio relevant clocking is done between the streamer and dac.
The data can’t be different at the source / destination or the built TCP in error checking would diagard the packet and a retransmit would occur. This all happens in milliseconds so a buffer of just a few seconds would compensate for any possible packet loss.
The only network requirements I ran I to to with critical timing requirements was for the trading companies; mega expensive switches designed to reduce latency at the sub millisecond level. The finance world loved this as milliseconds = $ when you are issuing trades.
1 Like
Just waiting for the dust to settle on this one…
Clock phase noise does not impact TCP. But, according to John Swenson, it does get embedded in the signaling and gets to the DAC. I don’t claim to understand the details…but that is what John Swenson proposes and says the EtherREGEN cleans up the noise. I just want to see something that shows the before and after of this.
FWIW, here is a link to the YouTube video in which a guy tests the ER and says it’s bogus. I don’t know if his test has any meaning, or if he knows what he’s talking about. But it is at least someone providing a test. It’s under 4 minutes.
It does seem a little odd to claim that clock phase noise effects the DAC negatively and have no data to support either the problem or the solution.
Perhaps @tedsmith can let us know if clock phase noise does indeed effect the DAC? Is this a known issue that DAC designers account for?
