Jump to content
IGNORED

UpTone Audio EtherREGEN Listening Impressions


Recommended Posts

4 minutes ago, skatbelt said:

My experiences with the EtherREGEN thus far:

 

Prior to it I already had musicians with me in the room so I expected nothing less than musicians embracing and hugging me with the EtherREGEN in place. Did this happen? The short answer is yes, somewhat, but certainly not out-of-the-box.

 

I came from:

Uptone modded Mac Mini as Roon server powered by Uptone JS-2 > copper ethernet (UTP) > Emo Systems EN-70HD network isolator > copper ethernet (UTP, short cable) > dCS Network Bridge > Chord DAVE > rest of the system.

 

Note that this setup was not physically connected to any router or switch. Streaming content and Roon control went over Wifi, setup in macOS via a bridged connection and internet sharing. This setup sounded very good and better than everything else I tried with (grounded) standard and audiophile switches. 

 

Another thing I have to mention is that the audio-part of my system is on a dedicated power line and behind an isolation transformer. So the EtherREGEN to me seemed THE solution to get complete (galvanic) isolation between the computer/network-part and audio-part without having (for instance) optical network components within the audio-part of my system.

 

So with the EtherREGEN I started with:

ISP-router > copper ethernet (STP) > EtherREGEN (powered by second rail of JS-2) A-side / B-side > copper ethernet (UTP) > dCS Network Bridge > rest of system.

+ Mac Mini Roon server powered by first rail of JS-2 > copper ethernet (UTP, short cable) > etherREGEN A-side.

 

I let the system settle for a day or two. Well, sort of, because I had lost connection issues with my Roon server (see the other thread that is quite derailing at the moment). This is resolved for now after disabling EEE.

 

After the two days I did some serious listening. And I found the results disappointing. Especially after all the positive stories I had read in this thread. I thought I heard a slight improvement in the microdetail department but the overall picture was less coherent, restless and no longer flowed the way I was used to. Beforehand I already expected the effect of the EtherREGEN to be not so impactful in my system because the dCS Network Bridge buffers the ethernet stream and does a lot of reclocking duties. One can literally unplug the ethernet cable and unit will play on for few seconds. But I was hoping the extra isolation relative to the Emo Systems isolator would at least bring some improvement.

 

When I had overcome this first setback I started experimenting with bringing that same Emo Systems isolator back into the chain. First between the ISP-router and the EtherREGEN. This seemed to be a logical choice because it would fight potential introduction of noise on the A side of the EtherREGEN. But against my expectation, this further deteriorated the sound quality. The same was true when I placed the Emo Systems isolator between the B-side of the EtherREGEN and the dCS Network Bridge.

 

But when I placed the Emo Systems isolator between my Mac Mini Roon server and the EtherREGEN (so also on the A side): eureka, everything fell back into place! As a result, the system now plays on a higher level on all aspects. Not in the superlative way others are experiencing but nonetheless.

 

What could be the (technical) reason behind this? Is the A-side of the EtherREGEN passing noise thru from the noisy ISP-router to the Roon server? Is the A-side of the EtherREGEN noisy itself and does it effect the performance of anything connected? Anyway, my strong advice - if you own an Emo Systems, Baaske or other type of passive network isolator - is to experiment with this yourself!

Great detail on the setup here @skatbelt, it helps a lot. 

 

The STP cable from the ISP router, most likely carries with it, unwanted noise, since it's coupled to a 'common ground' * directly, or by capacitance (hope not).

By fitting the EMO isolator, it's a transformer type based, so the isolator decouples the shield continuum from the rest of the system, much in the same way as the Jensen hum eliminators work for un/balanced line audio levels.

I would keep the EMO where it is, since it provides protection for high volts on Ethernet systems caused by destructive mains powered events.

Also possible that the STP formed a ground loop, perhaps can be overcome by connecting the ER ground terminal to earth (the same centre point from the IsoTransformer on the secondary) in this way all ground (earth) references are the same for the whole system.

 

The ground loop current or the unwanted noise "injection" is the main cause of loss of sparkle, micro detail and sound stage degradation. With placing the EMO in the right spot (for now), the missing attributes are now restored. This typically occurs with USB, which is difficult to control, but Ethernet can be corrected with simpler measures.

 

Please report back further listening observations.

 

* May not be connected to earth

AS Profile Equipment List        Say NO to MQA

Link to comment
6 hours ago, skatbelt said:

I agree on the fact that the ER is upperclass compared to Baaske or Emo systems isolators BUT in my system the end result is the best while keeping the Emo systems isolator between my Roon server (a modded Mac mini) and the ER on the A-side. So either it is keeping the A-side or the server cleaner (or both?).

 

I would not be surprised if the EtherRegen has a feedback sensing circuit which monitors noise, impedance levels, perhaps, between the B connector and the NIC in the digital audio device. It's like a load monitor if you like where the B connector source measures and compares to a set of parameters, where optimally noise is the lowest. Not sure how this is actually implemented, but John is clever enough to be aware that an open system has its drawbacks and would vary with the type of NIC used, traffic throughput and effects of reflections and the cable. 

 

From what users describe, the experience is much the same from person to person when connected directly (Connector B to NIC), other than adding in the Baaske or Emo, where the sound becomes worse.

 

When the Baaske or Emo transformers are installed between the B connector and the NIC in the load device, this isolates the load monitoring, and the correction circuits in the Etheregen 'max' out since it tries to correct for something that's missing. Consequently audible results would be apparent. There's no need to raise an alarm, since an open circuit on the B side can occur and shouldn't have any nuisance value.

 

As far as the A side is concerned, there wouldn't be any monitoring at all, no need to, really only look for for valid data, router signals much like a regular switch. When the Baaske or Emo is installed on the A side, the noise is reduced from the source due to the transformers built into these units. They also provide valuable overvoltage protection, which would be recommended in any case, no need to uninstall them!

AS Profile Equipment List        Say NO to MQA

Link to comment
  • 1 month later...

@flkin, thank you for the post and diligence in finding the best method to suit your ear. Am trying to find a reason why the treble is not the best. It is known from galvanically seperate devices like the iGalvanic, the treble is a victim of creating false noise as well as removing USB noise from the source.

Consider then, the hard treble is already present in the system, somehwere, let's start from the PF to the DAC. It's not present when the EtherRegen is installed since the effect is masked by the cables and to a certain degree common mode noise (aka leakage currents) from what's already being contributed by the other components. It's great to see star earthing, every bit helps.

When the EtherRegen is added, that path that allows the masking is gone, and you then hear the bad treble. To confirm, you could replace the Habst USB cable with a USB printer cable. If the treble is worse (and most likely all other sounds will be), then you've tracked a noise source.

I am really leaning toward that the EtherRegen is doing it's job and there's a problem in a section of the rest of the system and agree with @Guidof but more in the digital chain. The clue was no matter what PSU you used for the EtherRegen, the hard treble was always present, since the EtherRegen is agnostic to its power supply.

AS Profile Equipment List        Say NO to MQA

Link to comment
  • 1 year later...
6 hours ago, chungjh said:

I have Paul Hynes SRT4 LPS powering ER. I am using Lumin app and ER is connected directly to the wall ethernet. I don't have HF issues per se, just some sibilance only on certain female vocals like Diana Krall. I like the way my system sounds overall. With or without the ER, there is sibilance on some of Krall tracks. 

I must admit to hearing sibilant from Diana Kraal recordings. Since the Baaske isolator, Jcat net femto and bonding chassis and drives together I don't hear many cases , cause I normally stop listening and try something else. Lumin U1 for nearly two years now.

AS Profile Equipment List        Say NO to MQA

Link to comment
  • 1 month later...
3 hours ago, LowMidHigh said:

 

I received my 3rd ER a few days ago. As expected, during the first 3 days, the SQ took a step back, as it takes time for the ER to settle in. In the present, my categorical conclusion is that adding a third unit gives the listening experience a boost—although not as much as when adding a 2nd one. In my system, rough edges are smoothed out, and some details come to the fore more distinctly. The sounds stage is a tad stretched out too. Again, it’s not a dramatic change, and some people may find it not the best-bang-for-the-buck, but I’m perfectly happy. 

Perhaps report the findings in 11 days from now?

AS Profile Equipment List        Say NO to MQA

Link to comment
  • 8 months later...
On 2/17/2022 at 2:17 PM, watts said:

Okay, got about 200 hours on the etherregen now. Like many have claimed, a layer of grain has been removed.  I wouldn't characterize it as "huge", as far as upgrades and sound quality improvements go, as I have done quite a few lately, but the improvement is definitely there. (Perhaps my ethernet is not as dirty as others?)

 

A quick listen and I could see someone thinking it is like detail has been lost, but that was the glare and haze that was riding on top of the detail. But I didn't think there was any grain or glare in my system! Low level details are more easily identified, as the background seems a little blacker, and perhaps dynamics seem slightly increased as a result? Definitely smoother and more analog sounding. 

 

I'll add a few choices to the many good suggestions in this thread: 

 

Fleetwood Mac- Second hand News

3 tracks of guitar, the first one in the left channel, another in the right, and a muted low level strum in the right centre. Pick any of them: All three are better resolved and more natural with the ER in the chain. There is so much going on in many of these Lindsay Buckingham produced gems. 

 

 Neil Young - The Needle and the Damage Done

One guy and his guitar. What better way to show the analog sound. His voice sounds more realistic, and the guitar sounds, well, more acoustic. And of course every file is MQA free 44.1 Khz 16 bit: thanks for coming back to Tidal Neil 😍

 

Now to hook up this Audio Sensibility Supra ethernet between the ER and the Playpoint in place of the $2 one and see if that improves anything.

 

IMG_2204.thumb.JPG.25376536a864f5771cba510a02a6a00f.JPG

200 hours is a little too short. Give it 360 hours and the ER will open up even more. The sound will plateau after that and that grain will be gone. 24/7 burn in with signal even better.

AS Profile Equipment List        Say NO to MQA

Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



×
×
  • Create New...