So I've had an EtherRegen in rack for 2 weeks...

[plissken]

1 hour ago, davide256 said:

I'm going to give one a try since if it doesn't help, it does have a return period. After 20 years of supporting a tier 1 ISP and fortune 500 customers, I'm skeptical that anything at layer 2 internal to the switch can matter but am willing to consider that electrically connected switch ports might  degrade USB audio on susceptible endpoints with adequate performance potential. I've tried a Cisco switch before, made a marginal difference but not enough to justify adding "spaghetti" to the simple NUC with powerline Ethernet setup I have now.

 

I returned mine. Took the $640, spent $214, and have 10GB Optical on the server and my N3150 based Windows 10 Pro play back system.

 

Here's the iPerf and the N3150 is most certainly the bottleneck:

 

Connecting to host fs-ll-1, port 5201
[  4] local fe80::9c9f:b5ef:606:bb5e port 55844 connected to fe80::1fd:9bd:dc5c:f2f port 5201
[ ID] Interval           Transfer     Bandwidth
[  4]   0.00-1.00   sec   306 MBytes  2.56 Gbits/sec
[  4]   1.00-2.00   sec   292 MBytes  2.45 Gbits/sec
[  4]   2.00-3.00   sec   306 MBytes  2.56 Gbits/sec
[  4]   3.00-4.00   sec   307 MBytes  2.57 Gbits/sec
[  4]   4.00-5.00   sec   303 MBytes  2.54 Gbits/sec
[  4]   5.00-6.00   sec   304 MBytes  2.55 Gbits/sec
[  4]   6.00-7.00   sec   303 MBytes  2.54 Gbits/sec
[  4]   7.00-8.00   sec   301 MBytes  2.52 Gbits/sec
[  4]   8.00-9.00   sec   304 MBytes  2.55 Gbits/sec
[  4]   9.00-10.00  sec   306 MBytes  2.56 Gbits/sec
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval           Transfer     Bandwidth
[  4]   0.00-10.00  sec  2.96 GBytes  2.54 Gbits/sec                  sender
[  4]   0.00-10.00  sec  2.96 GBytes  2.54 Gbits/sec                  receiver

 

Still 3 times less expensive, I have true electrical isolation so no leakage currents to worry about with my 20 meter long moat, and I have 3000% more throughput.

 

[Superdad]

On 1/25/2020 at 5:53 PM, plissken said:

Here's my setup:

Playback Machine:

Dell Optiplex with Intel Dual Port Server NIC (Pro 1000)..

Port 1 goes to the B side EtherRegen and the other side goes to port 23 on my Aruba 2530

Port 2 goes direct to port 24 on the Aruba

 

Hi Mark:

So let me get this straight:

Your DAC-attached endpoint (with big SMPS) has a LAN card which has two Ethernet ports.

1) You attached a cable from one of those ports your 24-port Aruba switch (from which you also feed network connection to an ‘A’-side EtherREGEN port);

2) You attach the other port of your PC’s NIC to the ‘B’-side port of the EtherREGEN.

 

Result?:

You completely defeated the active differential isolation and clocking moat of the EtherREGEN, allowing both the common-mode leakage and the upstream jitter/phase-noise to modulate the ground plane of the ‘B’-side.

 

So it is clear that you ignored everything that was clearly spelled out in the User Guide for our product.

Plus your conventional Dell Optiplex with big SMPS is not in the least representative of what the majority of audiophiles are using as their DAC-attached endpoint (typically a linear supplied streamer/renderer quiet PC dedicated to audio).

 

In addition, you ran a whole bunch of data tests looking to find what our product does. You might as well asses the distortion of a power amp by measuring the AC current draw—or have your doctor give you a cholesterol screening by having you step on a weight scale. You were looking in the wrong place. And even your listening test was done with the main feature of the product defeated.

Epic.

[mansr]

8 minutes ago, Superdad said:

allowing both the common-mode leakage and the upstream jitter/phase-noise to modulate the ground plane

Do you have those catchphrases bound to a keyboard macro or something?

[plissken]

1 hour ago, Superdad said:

 

Hi Mark:

So let me get this straight:

Your DAC-attached endpoint (with big SMPS) has a LAN card which has two Ethernet ports.

1) You attached a cable from one of those ports your 24-port Aruba switch (from which you also feed network connection to an ‘A’-side EtherREGEN port);

2) You attach the other port of your PC’s NIC to the ‘B’-side port of the EtherREGEN.

 

Result?:

You completely defeated the active differential isolation and clocking moat of the EtherREGEN, allowing both the common-mode leakage and the upstream jitter/phase-noise to modulate the ground plane of the ‘B’-side.

 

So it is clear that you ignored everything that was clearly spelled out in the User Guide for our product.

Plus your conventional Dell Optiplex with big SMPS is not in the least representative of what the majority of audiophiles are using as their DAC-attached endpoint (typically a linear supplied streamer/renderer quiet PC dedicated to audio).

 

In addition, you ran a whole bunch of data tests looking to find what our product does. You might as well asses the distortion of a power amp by measuring the AC current draw—or have your doctor give you a cholesterol screening by having you step on a weight scale. You were looking in the wrong place. And even your listening test was done with the main feature of the product defeated.

Epic.

 

You didn't pay attention or even watch the video I posted. See attached PDF. I clearly setup for single path testing.  Again watch the video and you will see me remove one cable, then the other, then both. all without an interruption in music.

 

"Plus your conventional Dell Optiplex with big SMPS is not in the least representative of what the majority of audiophiles are using as their DAC-attached endpoint (typically a linear supplied streamer/renderer quiet PC dedicated to audio)."

 

If I've optimized my PC, and Switch, out the wazoo why do I need your product?

 

I clearly paid attention. If it's not too much of an ask please do the same.

 

I'm willing to provide this switch setup with a dual nic, optimized PC running JRiver of your choosing, in a rig of your choosing, with blinded participants of your choosing, with source material of your choosing.

 

The only thing you and others won't know is what cable, if any, is physically connected.

 

Lets see if each of us will eat our respective dog food. And to quote you:

 

You completely defeated the active differential isolation and clocking moat of the EtherREGEN, allowing both the common-mode leakage and the upstream jitter/phase-noise to modulate the ground plane of the ‘B’-side.

 

Lets see if you know when that isolation has been defeated.

 

Let me know in what ways I'm being unfair.

 

 

 

 

 

 

eR test setup.pdf

[plissken]

2 hours ago, Superdad said:

In addition, you ran a whole bunch of data tests looking to find what our product does.

 

It was requested that some throughput numbers be ran. So I ran them. They were fine and showed the eR not to introduce any bottlenecks. It basically provided wire speed.

[Ralf11]

"clocking moat" is another one, mansr

[mansr]

2 hours ago, Ralf11 said:

"clocking moat" is another one, mansr

Does the clocking moat have clocking crocodiles (clockodiles?) devouring any jitters attempting to swim across?

[Jud]

Just wanted to mention that I'm very pleased indeed with the subjective effect of my ER. I was somewhat skeptical and not at all expecting it to make the subjective difference it did.

 

Of course I realize this could very well be a purely subjective effect, since I've done no objective testing whatever. Yet I've never enjoyed my music more or wanted to listen so often. So I'm very happy.

[plissken]

26 minutes ago, Jud said:

Just wanted to mention that I'm very pleased indeed with the subjective effect of my ER. I was somewhat skeptical and not at all expecting it to make the subjective difference it did.

 

Of course I realize this could very well be a purely subjective effect, since I've done no objective testing whatever. Yet I've never enjoyed my music more or wanted to listen so often. So I'm very happy.

 

I wish the power of suggestion held that amount of sway for me.

 

[Jud]

4 hours ago, plissken said:

 

I wish the power of suggestion held that amount of sway for me.

 

 

Who knows, maybe it does.  

 

[plissken]

7 hours ago, Jud said:

 

Who knows, maybe it does.  

 

I don't discount it...

[davide256]

On 2/8/2020 at 2:41 PM, plissken said:

 

I returned mine. Took the $640, spent $214, and have 10GB Optical on the server and my N3150 based Windows 10 Pro play back system.

 

Here's the iPerf and the N3150 is most certainly the bottleneck:

 

Connecting to host fs-ll-1, port 5201
[  4] local fe80::9c9f:b5ef:606:bb5e port 55844 connected to fe80::1fd:9bd:dc5c:f2f port 5201
[ ID] Interval           Transfer     Bandwidth
[  4]   0.00-1.00   sec   306 MBytes  2.56 Gbits/sec
[  4]   1.00-2.00   sec   292 MBytes  2.45 Gbits/sec
[  4]   2.00-3.00   sec   306 MBytes  2.56 Gbits/sec
[  4]   3.00-4.00   sec   307 MBytes  2.57 Gbits/sec
[  4]   4.00-5.00   sec   303 MBytes  2.54 Gbits/sec
[  4]   5.00-6.00   sec   304 MBytes  2.55 Gbits/sec
[  4]   6.00-7.00   sec   303 MBytes  2.54 Gbits/sec
[  4]   7.00-8.00   sec   301 MBytes  2.52 Gbits/sec
[  4]   8.00-9.00   sec   304 MBytes  2.55 Gbits/sec
[  4]   9.00-10.00  sec   306 MBytes  2.56 Gbits/sec
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval           Transfer     Bandwidth
[  4]   0.00-10.00  sec  2.96 GBytes  2.54 Gbits/sec                  sender
[  4]   0.00-10.00  sec  2.96 GBytes  2.54 Gbits/sec                  receiver

 

Still 3 times less expensive, I have true electrical isolation so no leakage currents to worry about with my 20 meter long moat, and I have 3000% more throughput.

 

I guess the part of your testing I'm seeing as unidentified is power supply quality. I've not had any joy from DIY build  PC endpoints using off the shelf ATX power supplies

like Seasonic Gold and Rosewill fanless. While my current software contributes a lot, using quality DC power supplies

for the NUC and 5v USB to DAC are what kicked performance far past the microRendu that was my past "gold standard".

Whats your power solution for the N3150?

 

FWIW I found RPI 3 solutions to be hopelessly bad for low level signal resolution even with better power supplies. Still puzzled why the 7 and 8 series NUC's sound better for audio

than a device like the MR whose hardware was purpose built for audio.

[plissken]

2 minutes ago, davide256 said:

Whats your power solution for the N3150?

 

Just a normal ATX solution that didn't perform any worse than a Cary Audio streamer I owned.

 

Properly engineered DAC's and all that.

[Ralf11]

be sure the fuses are oriented properly to avoid screechiness

[m3lraaHnevetS]

My guess would be the noise level of your system is so high that you can not hear the difference when you pull either cable.

Although you should have heard a difference when pulling the Aruba switch cable, because that one is adding to the noise level while the eR is not.

Is that okay for an answer?

[plissken]

15 minutes ago, m3lraaHnevetS said:

My guess would be the noise level of your system is so high that you can not hear the difference when you pull either cable.

Although you should have heard a difference when pulling the Aruba switch cable, because that one is adding to the noise level while the eR is not.

Is that okay for an answer?

 

What about the sound when both cables were pulled? That is the fundamental question:

 

Are you playing off the cable or are you playing out of buffer?

 

I now have a system that Jriver can buffer an entire Album. I did it with a Leo Kottke album 'One Guitar, No Vocals' to test out. It's 554MB of data.

 

I hit play and right away went over and disabled the 10GBe interface. The DATA transferred so incredibly quick I can't mouse faster and right click and disable an interface in Windows. But I was able to listen to the Album in it's entirety.

[m3lraaHnevetS]

huh? one troll pulling the disagree card get's my reply deleted? that's effective trolling right there, respect 😉 

[The Computer Audiophile]

5 minutes ago, m3lraaHnevetS said:

huh? one troll pulling the disagree card get's my reply deleted? that's gay effective trolling right there, respect 😉 

Not sure what you mean by your last sentence but I’m not a fan and neither are others. Stop. 

[Jud]

6 minutes ago, m3lraaHnevetS said:

huh? one troll pulling the disagree card get's my reply deleted? that's gay effective trolling right there, respect 😉 

 

You might wanna take the bigotry somewhere else. This place is about having fun with audio.

[plissken]

Look. I built a purposefully extra large hammer in the form of a 10GBe network to have an honest conversation around over the wire playback.

 

Everyone is allowed their opinion. We aren't allowed our own facts.

 

What my setup is showing is just how absurdly easy it is to build some infrastructure that can get data off the wire as quick as possible.

 

For a track we are in the millisecond range now and then the network connection goes  twiddling it's thumbs.

 

Plus I have a very effective fiber moat that is 66 feet long. No EMI,RFI. Also fiber is lower wattage.

[m3lraaHnevetS]

5 minutes ago, The Computer Audiophile said:

Not sure what you mean by your last sentence but I’m not a fan and neither are others. Stop. 

I wonder why my reply was deleted

[The Computer Audiophile]

3 minutes ago, m3lraaHnevetS said:

I wonder why my reply was deleted

It was considered thread crapping. 

[Ralf11]

This place is about being stylish with audio.

[m3lraaHnevetS]

To me it is pretty obvious that the OP 'tested' the eR to prove too himself what he already knew.

So it is very unlikely that he will ever change his mind (unless maybe he get's to listen to a system that already has a pretty low noise floor so he can actually hear what the eR does).

Therefor I suggested to Davide to stop trying and add users like Mansr, plissken and Ralf to his ignore list as probably many others have done. Am I not allowed to say this?

I am asking because to me it seems that my earlier reply stating this same thing was deleted somehow.

And I will indeed stop now and listen to some music 😉

[The Computer Audiophile]

17 minutes ago, m3lraaHnevetS said:

To me it is pretty obvious that the OP 'tested' the eR to prove too himself what he already knew.

So it is very unlikely that he will ever change his mind (unless maybe he get's to listen to a system that already has a pretty low noise floor so he can actually hear what the eR does).

Therefor I suggested to Davide to stop trying and add users like Mansr, plissken and Ralf to his ignore list as probably many others have done. Am I not allowed to say this?

I am asking because to me it seems that my earlier reply stating this same thing was deleted somehow.

And I will indeed stop now and listen to some music 😉

Music listening is always encouraged. 
 

This is one where it works both ways. Those who listen to the eR and love it, hate when someone enters a thread to say they are delusional.  Your comments in this thread were the opposite but equivalent. 
 

I suggest taking a break from this topic.