UpTone Audio EtherREGEN (Objective Discussion Only)

[plissken]

https://www.ap.com/analyzers-accessories/apx-overview/jitter/

[plissken]

1 hour ago, kennyb123 said:

I was looking forward to this thread as I was interested in a objective take on this product.  Not surprisingly the negative remarks outweigh the substantive objective commentary.  If there more valuable objective assessments were shared here I might agree that it was wrong for Chris to split this off.  

 

What negative remarks? There has been some people that need correcting however.

Quote

A comment was made about Amir’s test equipment being able to properly capture jitter. Is that the case or not?

 

The comment was incorrect about the 600ps as it was conflated to the Analog inputs on the A.P. when it was a measurement for the Digital I/O on the AP which isn't being used.

 

Quote

 

 I realize that a video was shared but why not refute what @Superdad wrote directly?  I would find that to be an interesting discussion.

 

The video should have cleared some things up though. Also this is being refuted directly. Over at ASR.

Quote

I also wonder about the appropriate way to test the ER’s benefits at the output of a DAC.  What I’m not seeing is as a testable hypotheses put forth on how this would show up at the output.

Shouldn't UpTone provide this? They are making the claim.

 

J-Test is one such way. If we think a downstream change of input is going to effect end endpoint device output then we obviously measure the output.

 

Quote

This would have to be tested in a way that can detect the impact of jitter being reduced.  First introduce a known amount of jitter - take a measurement at the output of the DAC - then remove it - take another measurement.

 

This has already been done:

 

 

Quote

 

 

 If no difference shows up then adjust the test methodology until it does - then you are ready to test the ER.  Having said all that, this is exactly the kind of discussion I’d expect to see in an objective thread.  In other words, a discussion about devising the correct testing protocol to evaluate the claims made in John’s white paper.  

 

 

The way to test is to test the claims being made. The U.T white paper is a paper about jitter on Ethernet ports (totally disregarding the TX/RX buffers on the interface, ignoring the Host OS buffers, Ignoring application buffers) some how interfering with DAC output and corrupting that output. Output which is a voltage device. . This means the analog realm since it's obviously an analog signal at that point.

 

I'm glad you mentioned testable hypothesis. Since the manufacturer hasn't posited one, I created a video showing one way to do this. Now @Superdad did take a run at trying to discredit it by saying I had defeated the isolation. When I pointed out that I hadn't then it was to attack my proof of concept setup where again some things were again mistaken by Alex. So that begs the question: What setup can I put together that will show the efficacy of the eR. I.E. what setup (source, interconnects, amp, endpoint, speaker, source music) etc can I gather to replicate this on my own?

 

 

 

[plissken]

9 minutes ago, Jud said:

 

OK, one followup, and be gentle because I'm not technical so this is ignorant: If jitter on the digital side can be measured with an analog test, can jitter on the digital side of the test equipment affect analog jitter measurement sensitivity? Or is this completely separate and unrelated circuitry?

 

Hope this answers your question: Amir isn't using any of the digital portion of the AP. He's simply setting up DAC's as normal and plugging their outputs into the Analog input on the AP.

 

In the AP video the engineer mentions internal loopback so what that means to me is that it's all isolated and everything stands on it's own.

 

[plissken]

13 minutes ago, jabbr said:

My hardware is certified that this is irrelevant — either nonexistent or insignificant, but 1G or 100m hardware is not required to meet end to end jitter standards.

 

^^ I wish more people understood this.

 

[plissken]

1 hour ago, opus101 said:

There is the evidence of a considerable number of happy customers to consider - but with your ideological stance you prefer to dismiss that.

 

I would prefer to validate that. Claims without evidence are just as easily dismissed without evidence.

 

I would like to see what test rig UT developed against, or say at a show like RMAF or Axpona have UT provide a debiased listening session.

[plissken]

12 minutes ago, opus101 said:

 

What is 'UT'? I'm with you on the points about jitter, despite having read the white paper I still cannot see how jitter is an issue unless DAC designers aren't implementing best practice in terms of PCB layout. Which is why I'd like examples of DACs in the marketplace where this has been screwed up.

 

On the CM noise/isolation front, I would agree that fibre renders it a non-issue. Is anyone making a DAC which has fibre as its input?

 

UT = UpTone

 

Lumin makes one.  But don't miss the forest for the trees here: If copper RJE connectivity is so problematic, why does Auralic, Lumin, NAIM, Cary Audio, Cambridge etc, etc... all have that type of connectivity?

 

Given the expensive measures I've seen many go through optical based Ethernet is kids play.

[plissken]

11 minutes ago, opus101 said:

 

How would you validate it? Jud saying he's a happy punter should just be dismissed as 'no evidence' ? If not, then what?

 

SBT testing would be one way. Capturing some output would be another.

 

Jud posted two files where he was having some form of ground loop induced noise. I listened to them blind and could 100% pick out the problem child.

[plissken]

1 minute ago, opus101 said:

 

Are you curious as to the reasons those companies make DACs with copper connectivity or are you arguing to make a point like 'Those guys make DACs with those interfaces so cannot be so problematic' ? If the former, have you asked them? If the latter, consider how many manufacturers make products with single-ended analog inputs and outputs using RCAs.

 

If copper based RJE is such an issue, and these are outfits with smart engineering staff... It's simply a question I believe with 1 of 2 answers:

 

1. Copper RJE isn't a problem

2. As a luxury item with appreciable costs for a lot of people I would hope the best possible interface technology to be right there with the rest of their products pedigree.

[plissken]

15 minutes ago, manueljenkin said:

Umm. So, a published, reviewed paper will do? Let me know the list of things you are 100% certain of! Fr is all you need to know in a headphone? All DACs are the same? What else?

 

I'm certain properly implemented Copper RJE will deliver stellar results in an audio system that can not be improved upon with an audiophile switch.

 

As I've mentioned I'm willing to pay out to be disabused of the notion. Anyone can PM me for details.

[plissken]

3 minutes ago, manueljenkin said:

Adding to that. We have things like packet loss etc. I've heard real packet loss, buffer underrun, etc by tweaking my buffer settings keeping the rest of the things bit-perfect. You don't necessarily need to hear drastic clicks or pops (though I was able to make it worse enough for that scenario also to happen). The lesser ones just get concealed (delta Sigma dac acts as a natural packet loss concealer).

 

Are you talking about USB packets or IP?

[plissken]

1 hour ago, Superdad said:

Precisely which electrical measurement characteristics do YOU think will correlate to the differences the reviewer hears between all the elements in his highly refined digital chain?  Do you believe that the “ASR verified” under $1,000 system you put together (and spoke about in another thread here) with $400 Yamaha integrated amp/streamer (a nice desktop piece though) will present music as compellingly and accurately as the one used in this review?

 

In one of your other posts—also in reply to a product review article on this site—you wrote:

 

”A performance analysis with a lab grade audio analyzer showing actual differences in output, or it only happened in your brain. That'll be the only evidence anyone with a extensive science and engineering background will ever accept. Our senses are too unreliable, too easily fooled by our emotion states and perceptions, to be accepted as valid evidence. So, yes... my mind is closed to opinions and anecdotal "evidence"... ”

 

There are other threads on this forum where you can take that debate. But it is not likely anyone reading this review is going to take your bait... 

 

[EDIT: Note that I recall writing the above reply when Mr. Bill’s post was made in the EtherREGEN review thread. I guess Chris moved both posts over to this thread—which is clearly where such debate should occur if desired. ]

 

How can you knock someone else's system yet at the same time, as of date, not provide a system that this can be demonstratively proved on?

 

What DAC/Computer or network streamer and switch can we use with your product do perform quantitative testing (either instrumented or truly ears only) with?

 

So far the only empirical evidence is what Amir has produced. Amir also produced evidence that your USB Regen can make a difference on a singular Schiit product and he also proved, and you entrenched on this one but finally capitulated in the face of measurements, that your meanwell smps was defeating the LPS isolation.

[plissken]

In a bit I will be posting some ADC captures of audio playing from a $400 audiophile switch, a $19.99 D-Link GO-SW-8GE, and no connectivity what so ever.

 

Since the white paper premise is that we are affecting the DAC analog out and we are thus capturing that output....

 

Everyone will be able to listen and decide when it was an Audiophile Switch, a $20 D-link, or no switch at all.

 

I'm up for track suggestions and if people want to make a 16/44.1 track available just PM a D/L link. This should meet 100% the definition of objective. I'll capture a screencast of it.

[plissken]

2 hours ago, Superdad said:

 

That's pretty funny. You do realize that your ADC (unless it is some wildly spectacular unit) is going to wipe out any differences to be heard.  How about you first use that ADC to capture audio from a $100 DAC and a $5,000 DAC and have people report on what they hear. 

 

It's not funny because you posted about the differences being on the output. How does the inputs on an ADC wipe out differences but the inputs on an Amp won't?

 

Again what DAC did you test this on?  I'll see if I can't get a hold of one.

 

 

[plissken]

1 hour ago, Superdad said:

 

We did.

But some do not seem happy with that.

 

 

Happy with what? I read the attempted FUD (Fear Uncertainty and Doubt) paper. One that reading the FAQ simply tells people to :

 

1. Go optical to break leakage current

2. Get as much wire speed as humanly possible with a playback option that caches the entire track up front so we only have to deal with phase noise for the 30-40ms that it takes to cache the track.

 

[plissken]

Sandy, read Alex's FAQ in the 'white paper'. It's exactly what people are being told to do as best practice.

 

In regards to phase noise the eR is actually working against an audiophiles interest by forcing the connection to 100mbps. That 100mpbs ensures that you are spending more time generating phase noise.

[plissken]

Sandy.... I wasn't speaking to you. I was speaking to Alex about the White paper and what UpTone is really suggesting people do.

 

You disagreed with it. I pointed out two sections of their 'paper' that said:

 

Optical breaks leakage currents, data transfers generate 'phase noise'.

 

So go optical and go for as a high speed interface as you can until something else in the setup becomes the bottle neck.

 

In my setup it's a $69 Celeron 3150 based mainboard. But at 332MB/s I don't think it's detrimental. Would a J5005 based system get me closer? I don't know because my SSD may start being the limiting factor. Then I have to start chasing NVMe... And the cycle of insanity over processing a measly 706Kbps continues.

[plissken]

On the post you are trying to refer me to:

 

It's a reply to Alex, not you. You chimed in with your typical disagree button like a lab rat hitting a bell for cheese.

 

If you have a directly related point in this 'Objective Section' please make it or shut up.

[plissken]

Just now, The Computer Audiophile said:

$69! Preposterous! You could’ve spent half that and done better. Someone call the snake oil police!

 

I'm talking about the objective wire speed performance of $69 passively cooled mainboard. iPerf was used so it's all chunked out of RAM.

 

Maybe I could get faster wirespeed with a Opteron 6276 as you can get both a 24 core CPU and Mainboard for ~$50-60. Can't speak for the fan noise however...

 

 

[plissken]

52 minutes ago, The Computer Audiophile said:

$69! Preposterous!

 

Don't know: $640 for a 100Mbit switch which does nothing that a $40 switch can't do or $210 for 10GBe optical and $430 on a passively cooled, zero mechanical noise, computer system (J4105, 8GB RAM, 240GB SSD, JRiver, Monitor).

 

Implement some SoX, Room Correction, House Curve in JRiver...

 

I'll take my chances on what constitutes preposterous in most peoples mind.

 

 

[plissken]

2 minutes ago, The Computer Audiophile said:

I guess my joke didn’t go over well. 
 

I was joking that anytime you guys bring up how much money people could spend to get the same performance (and give a self satisfied sniff because it’s seemingly cheap), there is always another product cheaper than the one you mentioned. I’m not a fan of racing to the bottom though. 

 

I'm for what's effective. Cheap/Inexpensive are subjective terms that don't have any meaning on their own. Price to performance is another matter.

 

I picked the 3150N years ago because it met my primary needs: Passively cooled, PCI-e expansion, and more than enough processing power for running Convolution and 24/192. $69 was a side benefit.

[plissken]

Just now, sandyk said:

 

 

 DBTs are frequently demanded by Objectivists. Are a tool used by Objectivists, and rarely by Subjectivists.

 

Right now I would settle for either DBT/SBT or some reproducible measurement data.

[plissken]

1 hour ago, sandyk said:

 

Are you willing to cover the costs of having this done independently from A.S. and A.S.R ? 

 

What I've proposed is working out the mechanics (I posted a YT video to this setup) and allowing for comment as to any shortcomings. The only comments were by Alex and he simply didn't have the correct understanding but we worked through that. I think.

 

So I guess what is your definition of 'independently' in it's entirety. Some things like 802.3ad are simply standards and any vendor hardware that supports it should suffice. Be it lowly TP-Link, Netgear or HPE Aruba, Cisco, Juniper, Arista...

 

I would rather work with golden ear claimants instead of random people off the street.

[plissken]

4 hours ago, The Computer Audiophile said:

With all due respect, I'd think an objective leaning person such as yourself would do a bit more homework on his sources of information. Check if any of the products Amir sells are "reviewed" on his site? It wouldn't do well for his pocket book if he had to tell his customers that the stuff they are buying from him doesn't measure as well as the cheap stuff. 

 

Way off the mark there Chris. Look at the Elac  DBR-62 (he doesn't sell) vs the Revel M16 (he sells).

 

He doesn't sell any Matrix or RME but they are his suggested buys among others.

 

It'll look better for you if you come off your Amir hate and spreading these false allegations.

[plissken]

I have a question: How many anecdotal reviews does it take to reach factual basis when we have factual evidence that runs counter?

 

Do I trust a $28,000 analyzer that's entire pedigree is analog measurement in the human hearing band, or sighted bias?

[plissken]

3 hours ago, Superdad said:

 

Obviously you trust the former--to tell you that the latest $125 Topping miniature DAC sounds just as good as $10K-$30K reference DACs. That's what we see talked about over at that "other" place. And according to Amir, the mighty Topping must be vastly better than your $300 Emotiva Stealth. Guess it's time for an upgrade, eh Mark? 

 

With BT being integrated with some of these DAC's it is about time for me to upgrade.

 

It's not only the Topping that has surpassed my DC-1. There's quite a few under $1000 DAC's I'd put up against equally well measuring $10-$30K reference DAC's.

 

Let me know if you want to blind A/B at Axpona in August. We'll get it all unity gained and go from there.