- Joined
- Feb 28, 2016
- Messages
- 1,279
- Likes
- 1,180
I'm going to get a wall of o-scopes behind me so I will be X times smarter than Hans.
-
WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
UpTone Audio EtherREGEN Switch Review
- Thread starter amirm
- Start date
I'm going to get a wall of o-scopes behind me so I will be X times smarter than Hans.
Add a Jacob's Ladder and a giant open 3 pole knife switch and you get to be an honorary mad scientist.*
(A real Mad Scientist requires a 4 year degree in Mad Science, a 5 year internship under a licensed Senior Mad Scientist, and a passing grade on all 4 stages of the Exam. And at least one minion.)
He pulled mine too
AudioEnthusiast
Member
- Joined
- Feb 8, 2020
- Messages
- 55
- Likes
- 44
I agree. I unsubscribed from his YouTube channel recently. His descriptive reviews are useful, but recently he has gone off the deep end. I felt bad about it. His channel explained file based audio and introduced me to Roon, for which I’ll always be grateful, but the kookiness is more than I can take.
but but every audiophile loves itThis is a review and detailed measurement of UpTone Audio's Ethernet "audiophile" switch. It is on kind loan from a member. The EtherRegen costs US $640.
The EtherRegen as the name indicates, "regenerates" and Ethernet signal. This is a functionality that is provided with any Ethernet switch that exists in the market. There are however claims of improvement which I will get into later. For now, the EtherREGEN comes in the same small aluminum enclosure rest of UpTone products come in:
Even though this seems to be the "back," it is the business end of the unit in that it has one special port ("B") that is supposed to be isolated from the ports on the other side:
The power supply is private branded but I am assuming that is just a sticker and UpTone has not really applied for FCC and regulator certification itself. This always puzzles me as companies like UpTone make a lot out of noise in audio yet they don't get their products (as required by law) certified for emissions per FCC or CE. For all you know the EtherRegen may be creating a ton of more high frequency noise ("EMI") than a cheap but certified Ethernet switch.
Note that the AC cable to the power supply is quite short. So you may have to replace that to have better reach.
Anyway, back to the audible claims, we read such on UpTone's website:
Starting at the bottom, data integrity is not an issue whatsoever in home LAN networks as evidenced by the fact that you are having no problem using your networked devices without this box. I monitor my 48 port switch form time to time and I have yet to find a single error on any of its ports.
Since Ethernet is designed to go long distances (100 meters/330 feet) by design it has to tolerate differential in ground potential, noise, etc. As such claims of better performance here -- without measurements to show the same -- is immaterial.
Perhaps the biggest issue with claims of audio improvement is that your DAC is so far removed from Ethernet that little you can do upstream can impact it. Ethernet has a clock but that is used for communication on the wire. Once a packet (chunk) of data arrives, it is put in memory in the operating system. At that point, it no longer has any timing information much less a clock. It is the responsibility of the application to associate timing with it. And such software notion either works, or doesn't. If it doesn't your music will stop or drop out. None of that timing has any relationship whatsoever with the clock that the DAC eventually uses to play data sent to it. It is the audio application together with the DAC (and or Operating System) which determine timing.
To wit, you can transfer said networked data to a memory card and put it in a player that has such a card, and play it. Surely at that point, there is no sense of timing as the device is not even reading from the network then.
Just in case someone doubts my qualifications on any of this, I worked on one of the first Ethernet interfaces back in 1984 or so which cost some $25,000 (?) and plugged into a $300,000 minicomputer. I worked on the TCP/IP protocol in Unix operating system that interfaced with that which powers the entire Internet and your home network. I also worked for two companies that produced end to end encoding, streaming and playback of AV content on the Internet (a start up which then got acquired by Microsoft). I can go down to even lines of code if you like to describe what is happening here.
Problem withe above argument is that you have to take my word over that of UpTone folks and that may not be an easy decision. So we resort to objective measurements to see if addition of EtherRegen changes the sound waves coming out of a networked DAC.
As luck would have it, I just received the Matrix Audio's Matrix i streaming DAC. It retails for US $990 and prior Matrix products have provided exemplary performance. So we know it is not the "limitation" in such tests. I could use my PC for testing but there is no such going on in a PC that there is no hope or prayer that any noise from an Ethernet switch would make a difference. In other words, I am doing what I can to help the EtherRegen do its thing.
EtherRegen Audio Measurements
I powered on the Matrix i and selected Network input. I connected its balanced XLR output to my analyzer for the tests you are about to see (at the end, I also tested RCA but made no difference). First our usual dashboard of 1 kHz with just my cheap, generic TP-LINK gigabit switch feeding the Matrix i via Roon streaming:
View attachment 41430
Our highest distortion product is at less than -120 dB, assuring 100% transparency (best case dynamic range of human ear is 116 dB).
I disconnected the cable from my generic switch and used the same cable to feed port A on EtherRegen. I the used one of the nicer, short Ethernet cables I have to go from A side of EtherRegen to Matrix i:
View attachment 41431
Even though there are usually run to run variations in such a measurement, it is remarkable how stable both the Matrix i and my Audio Precision APx555 analyzer are. Output voltage and frequency are essentially dead on. Distortion+noise is varying by a fraction of a dB.
Next I connected the Matrix i using the Port B of EtherRegen:
View attachment 41432
Nothing changes.
Words like "jitter" are used in UpTone marketing so perhaps the improvement lies there. So let's run that test but go nuts by increasing FFT length to whopping 1 million points:
View attachment 41434
You see a difference? I sure as heck don't.
There is actually a bit of jitter around our main tone of 12 kHz. But EtherRegen does nothing to reduce that either. A few odds and ends are there too but at levels below -140 dB, absolutely not a concern audibly.
Let's confirm port A and B on Ether Regen:
View attachment 41435
Nothing there either.
Maybe the goodness comes at higher frequencies than 20 kHz:
View attachment 41436
Once again, Ether Regen provides no improvements whatsoever. It is as if it is not there which is what the architecture suggested all along.
We could stop here but why not turn some other stones. "Noise" is another thing audiophiles worry about. So let's run the same test as above but this time, with nothing playing:
View attachment 41437
Whether we look at this up to 20 kHz or 90 kHz, there is absolutely no difference in the noise output of the Element i with generic switch or either port on EtherRegen.
Let's go really crazy and run this measurement to 1 Megahertz:
View attachment 41438
All three graphs land on top of each other (last graph was yellow so dominates in color).
Thermal Measurements
Half-way through the test, I start to smell what resembled like overheating electronics. Decades of repairing electronics has sensitized my nose to such things.I get my nose close and realize it is the Ether Regen that is cooking! You could easily small it through the ports. While I still could hold it, it was running quite warm. I took an IR scan of it:
View attachment 41439
Case temp is between 36 and 40 degrees C. I worry how hot the internal components are running and whether the junction max temp is too close for comfort.
AC Power measurements showed that the switch is consuming 7 watts. That may not sound like much but for everyday electronics, it is good bit of power. We will have to see in a few years whether these things fail or not. If my math is right, running EtherRegen will cost you $9/year at 15 cents a killowatt-hour electricity cost.
Listening Tests
The beautify of how streaming audio works allows us for a very clever AB test. Both the host and client have fair bit of memory that they store Ethernet content into. With audio playback consumer data at a rather low rate, you can actually switch Ethernet cables and be able to play the music uninterrupted!
I connected one wire to Matrix i Ethernet port from my cheap switch, and another to Ether Regen.
So I started with my generic switch connection and started to stream some of my reference quality content. While the music was playing, I switched Ethernet cables, this time having Port B of EtherRegen feeding the Matrix i. There was absolutely no audible difference. I then switched back to generic switch feed. No difference. I switched back to EtherRegen. No difference. None. Nothing.
Conclusions
Computer, networking and streaming architecture instructs us that an external switch cannot have any effects on an Audio DAC. But it is always good to put some hard data behind this. And that is what the measurements show. That no matter how deep we dig down in the waveforms coming out of the DAC, no difference exists between a cheap, generic switch and UpTone EtherRegen. Even when going down to incredible -160 dB which is equivalent to a 27 bit audio word (3 bit more than any 24 bit content), there is still no difference.
Measurements conclusively demonstrate that EtherRegen did not change jitter, noise or distortion of the DAC. It further had no impact on its clock speed, or output voltage.
What is that? You hear otherwise? Well, I tested it and there is no audible difference either. I suggest you repeat my test. If you do hear a difference, have someone switch cables behind your back. You can do this quickly, or wait days or weeks to switch. Just don't look and keep a log of 10 trials. If you can tell better than 8 out of 10 times that there is an audible difference with EtherRegen, then you have conquered the impossible!
Now, I am not beyond paying more for something if it looks better, feels better, etc. Unless you have an upset stomach and like to use the EtherRegen as a hot water bottle replacement (don't laugh, I have done that with my laptop!), there is no such benefit. Even if there were, I would not pay $640 for it. $100 maybe. But $640?
Needless to say, I absolutely cannot recommend the UpTone EtherRegen. It seems to be designed by people who a) don't understand the architecture of streaming audio and hardware and b) didn't bother make measurements of said flaws before going to fix them. A problem was imagined, and then supposedly fixed.
--------
As always, questions, comments, corrections, etc. are welcome.
Having someone over in person on Friday to make measurements. Thought I should get some food to be polite but am too cheap to spend my money on it. So would you donate a few dollars? Remember, I have expensive taste in foot so be generous: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/
I don't
OP
- Thread Starter
- #587
I am an audiophile but don't love it.
- Joined
- Aug 6, 2017
- Messages
- 149
- Likes
- 91
Pretty much. Over 1,100 units sold and shipped and only about 4 returned under our 30-day, money-back guarantee.
Another long review out today:
https://audiophilestyle.com/ca/reviews/uptone-audio-etherregen-review-and-comparison-r887/
as well as the release of our 'white paper':
https://audiophilestyle.com/forums/...f-the-uptone-jswenson-etherregen-white-paper/
Soundstage
Active Member
- Joined
- Jan 10, 2020
- Messages
- 294
- Likes
- 216
This is ASR. Bad URLs are not relevant nor white papers without a single figure. Reminds me of the Bitcoin bubble when new crypto currencies where popping out every week with the same language.
Shorty
Active Member
- Joined
- Sep 25, 2019
- Messages
- 182
- Likes
- 293
Plus a Wurlitzer.
Plus a Wurlitzer.
Maybe where you come from.
Happy customers are fine, but not valid proof that your product does anything of value. Actual numbers, like the ones Amir used to prove your product does not add value, would be needed. So show us some actual performance data, please. Not a "how it could possibly work" white paper, that also has no real value.
I guess my setup where I listen to my FLAC based collection over WiFi would be unbearably bad according to these frauds (maybe I need one of those atmosphere towers from Synergistic Research to alleviate it?). I've noted that the WiFi connection matters so little for the audio quality that I can even reboot my access point while listening to my music, without the music being interrupted. Of course, this would make sense if one had any idea of how computers and networks work on a basic level.
Pro tip, they do not function the same direct way as a record player does.
When things are so objective as they are in the digital space, there ought to be customer protection against companies that outright lie. On the other hand, showing this to my colleagues at work gave us a hearty laugh.
Pro tip, they do not function the same direct way as a record player does.
When things are so objective as they are in the digital space, there ought to be customer protection against companies that outright lie. On the other hand, showing this to my colleagues at work gave us a hearty laugh.
I'm still perplexed by how a product is developed without any measurements. How can one decide what is needed if you really do not have an objective clue of what is even going on before, during and after?
I suppose the 1100 units (and counting) @ $640 each = $704,000 is what drives the "design".
I suppose the 1100 units (and counting) @ $640 each = $704,000 is what drives the "design".
I'm still perplexed by how a product is developed without any measurements. How can one decide what is needed if you really do not have an objective clue of what is even going on before, during and after?
I suppose the 1100 units (and counting) @ $640 each = $704,000 is what drives the "design".
Copy circuits from datasheets or existing designs. Do some basic calculations. Maybe even a simulation, but I have my doubts. Then breadboard it up and swap parts until you like it. Do a layout, get back some boards, confirm they sound how you like. I'm sure they use an o-scope to confirm stuff, but those are not to be trusted for accurate measurements if you look closely at their spec sheets. Tweak a bit more. Have the golden ears squad give it the thumbs up. Send it out. You'll have something that works, but not necessarily works well, and I would be concerned about the reliability and stability of the product. But that's just me.
Just follow this simple plan.
Phase 1: Collect
Phase 2: ?
Phase 3: Profit
- Joined
- Feb 28, 2016
- Messages
- 1,279
- Likes
- 1,180
Oh dear lord. That white paper.
1: I am waiting with baited breath to see a 1Khz input and then see the the 10hz spuria (and the ability for it to be independently reproduced and also showing up in controlled listening.
2: Thanks for the FAQ as I've been doing two things: Caching the entire track and using 10GBe fiber. I can cache an 75MB track in about 300-400ms (about the speed of an eye blink). I did everything for $210 as a proof of concept. Less than 1/3 the cost and 3000% more bandwidth and more isolation than your 'product' can provide.
3: I see no mention that wireless solves all these problems. My $100 in dedicated WAP and WiFi adapter get me on average 38MB/s. So more than 1/6th less expensive and almost 4X faster on average.
I've seen the latest infomercial at Audiophile style with AustinPop leading the circle jerk.
Last edited:
OP
- Thread Starter
- #599
It is a well written paper about how jitter can be created. Problem is, it is just theory. If you measure as I did for review of your product, you see that the problems do not exist in any quality DAC:
This analysis shows a noise floor of -160 dB. Translating this into jitter components, it would mean that any jitter higher than just 5 picoseconds would show up above. The paper talks about nanoseconds. I am talking picoseconds that are 1000 times smaller. Converted to bits, the above noise floor allows us to see jitter components that are down at 27 bit resolution!!! Yes, if there is slightest jitter even in a 24 bit signal, we easily see it as is the case in the above display.
Translation, we have an extremely, extremely sensitive test here, way beyond anyone's hearing threshold. Yet, the Etherregen fails to show any improvement whatsoever.
Why is that? Well, everything John talks about in the paper and then some is already occurring inside any PC. Huge number of digital circuits are running around in a computer, polluting the ground on USB connection. An excellent DAC filters and/or makes the variations immaterial. Once there, you have no prayer of changing the DACs behavior one step removed from a DAC in an Ethernet switch.
Your problem is that you have picked an objective figure -- jitter -- which we can easily measure. Heck, just a good audio interface for a few hundred dollars would prove the same point I have. The fact that you don't measure or measure but can't find an improvement doesn't bode well for you.
Conclusion
John's paper points to a truism in system design. What it misses is that the problem he thinks exists, is already solved. A DAC has to deal with many sources of jitter already, many of which are far more significant than what the PC can produce, let alone an Ethernet switch feeding a computer. This is an open and shut case based on the very theories John puts forward. Claim was made, but did not translate to reduction of any jitter in the analog output of the DAC.
As you know, similar claims were made about your other products and the measurements proved them to be fallacies just as well.
Happy to have created some traffic for you and AS website. Hopefully in return people find our answers after searching and realize the truth that perhaps neither you nor John understand (hate to go to the place where you do understand but say differently).
Buzz kill!It is a well written paper about how jitter can be created. Problem is, it is just theory. If you measure as I did for review of your product, you see that the problems do not exist in any quality DAC:
This analysis shows a noise floor of -160 dB. Translating this into jitter components, it would mean that any jitter higher than just 5 picoseconds would show up above. The paper talks about nanoseconds. I am talking picoseconds that are 1000 times smaller. Converted to bits, the above noise floor allows us to see jitter components that are down at 27 bit resolution!!! Yes, if there is slightest jitter even in a 24 bit signal, we easily see it as is the case in the above display.
Translation, we have an extremely, extremely sensitive test here, way beyond anyone's hearing threshold. Yet, the Etherregen fails to show any improvement whatsoever.
Why is that? Well, everything John talks about in the paper and then some is already occurring inside any PC. Huge number of digital circuits are running around in a computer, polluting the ground on USB connection. An excellent DAC filters and/or makes the variations immaterial. Once there, you have no prayer of changing the DACs behavior one step removed from a DAC in an Ethernet switch.
Your problem is that you have picked an objective figure -- jitter -- which we can easily measure. Heck, just a good audio interface for a few hundred dollars would prove the same point I have. The fact that you don't measure or measure but can't find an improvement doesn't bode well for you.
Conclusion
John's paper points to a truism in system design. What it misses is that the problem he thinks exists, is already solved. A DAC has to deal with many sources of jitter already, many of which are far more significant than what the PC can produce, let alone an Ethernet switch feeding a computer. This is an open and shut case based on the very theories John puts forward. Claim was made, but did not translate to reduction of any jitter in the analog output of the DAC.
As you know, similar claims were made about your other products and the measurements proved them to be fallacies just as well.
Happy to have created some traffic for you and AS website. Hopefully in return people find our answers after searching and realize the truth that perhaps neither you nor John understand (hate to go to the place where you do understand but say differently).
Similar threads
