Objective proof the UpTone Regen ISO can improve a DAC's output(*)

[firedog]

 

If a product isn't what I see as ridiculously expensive and sounds like the idea behind it is reasonable, I'll give it a try and see if I think it makes a difference. In the end, there are endless sources of confirmation bias - some plus and some negative - for any product, so if I "think" something sounds better - then it does. 

 

Objective truth would be good, but I don't see how we arrive at it in practical terms. Very few products are tested objectively, and even fewer by more than one person. We also often have dueling "experts" telling us why something can or can't make difference. Since I'm not an expert, I basically don't believe anyone.

 

In Amir's original testing of the Regen, he made a basic mistake, got results which even he said didn't make sense, but went ahead and published them anyway. He soon corrected them, but his "rush to publish" makes me think that he may not be an objective source - even if consciously he is trying to be. I don't see an indication that he is purposely trying to get negative results for these products, but I'm not at all sure he doesn't have a bias which affects how he tests and his results/conclusions.

 

If you read iFi's long response at their sponsored forum to his testing using the iFi iPower, they give a pretty convincing argument that just having an AP analyzer and an engineering degree doesn't show that someone is competent to use it properly. Basically, they said that Amir makes mistakes in his setup that give him false results. And he isn't aware of them, because "you don't know what you don't know", as it were. There were a couple of others on the net who tested the iFi and didn't agree with Amir's results. 

 

I'm using an IsoRegen with a galvanically isolated USB input. The manufacturer didn't think a unit like the IR would help it (but they all say that). Alex gave an honest answer and said that it could still possibly make an improvement to this kind of setup because it would apply the isolation before the USB input and not after it  (internally) - as happens in most DACs. That was an honest response, not hype. So I ordered one. So far I'm not hearing a big improvement with it. But I don't have any method for doing unsighted A/B testing, and so my impression in itself could be due to some kind of confirmation bias. I'm going to keep testing my system with and without it, and at some point will come to a conclusion.

 

If I decide it isn't doing much, I'll sell it. I'll take a small loss. But I will have learned something and hey, it's all part of the hobby. We aren't talking about saving starving children here. Sometimes (and especially over at Amir's forum) it seems that is forgotten. 

 

 

[firedog]

2 minutes ago, scan80269 said:

A battery costs much less than an LPS-1 and should serve just as well to break an AC leakage loop.  The LPS-1 is not the only solution, though it is a premium solution that brings some great specs besides AC leakage firewalling.

 

Measurable differences are not always audible (e.g. the bumps of -110dBFS 60Hz & harmonics shown in Amir's graphs with ISO-REGEN - confirmed by several here on this thread alone), and audible differences are not always demonstrable with measurements.  In other words, our measurement capabilities & methods (e.g. AP & FFT) are currently not advanced enough to reveal everything our ears can discern. 

 

To insist on seeing some difference in some measurement as the only criterion for the merit of a product is one of Amir's biggest fallacies.  He is an audio scientist or "specophile", not an audiophile.  Audiophiles place their faith on the most sophisticated audio instruments available: their own ears.

 

Great engineering plus critical listening are essential to delivering good sound, but great specs? not so much.

 

All of that is sort of irrelevant. If Amir's tests are properly done, then he has a good point-the the IR isn't doing what it claims to -or at least not all of what it claims to. The idea that he should be using an LPS-1 to test the IR makes no sense, as Uptone sells the device with the Meanwell and nowhere claims that the IR won't do what they it is supposed to do when powered by the MW. 

 

[firedog]

1 hour ago, scan80269 said:

One of the disputes here is whether Amir's graphs are representative of what the ISO REGEN is supposed to do, as claimed by its manufacturer.

 

Being a USB regenerator, ISO REGEN is intended to clean up the USB signal going from source to DAC.  From measurements perspective, the measurements should be focused on quantifying the USB signal quality differences at both ends of the device.  In this regard, I have to say the eye diagrams UpTone posted for USB signals entering and exiting an ISO REGEN are highly relevant.  Crappy measured USB signal coming from source (entering ISO REGEN) converted to clean measured USB signal (exiting ISO REGEN) going to the DAC.

 

Amir's work & conclusions on ISO REGEN are not unlike doing a smog check on a new car, then using the measured results to refute the car manufacturer's claim of engine fuel efficiency.

 

Uptone specifically claims that the USB cleaning done by the IR "makes the music sound better", not just that the USB output of the Regen is cleaner than what went in. 

 

If such devices make the music sound better, the only way they can be doing so is by resulting in a better sounding output from a USB DAC, than the output from the same DAC with no IR in the chain. If the output of the IR is cleaner, but this output has no effect on the the output of the DAC, then what is the point of the IR?

 

So this would logically need to be measured at the DAC, not at the Regen. 

 

Then the question becomes, do we know what to measure at the DAC and are we capable of measuring it? And secondarily, is Amir measuring the right thing and does he know how to measure it properly?

 

[firedog]

4 hours ago, Sal1950 said:

Come up Chris, seriously. When folks run around making claims in any endeavor they are expected to be able to back those claims up.

When they are then offered the opportunity to do so, and show the whole world they are right, and refuse to even participate, there's no reason to think hard at why, or attempt read things into it. The actions speaks for themselves.

There is a difference between responding to a manufacturer who makes a claim about a product and a user who posts that the device made his system sound better. YOU may think you have scientific proof showing the device can't do that, but actually you only have a "science-like" indication. 

 

Nothing has risen to the level of proof for how something sounds in another setup.

 

So, IMO, there is little point to respond to a specific user that his ears are fooling him. You actually don't know that. 

 

In addition, home users have basically little or no ability to conduct anything like a scientific or truly blind test on a device. So they audition as best they can and make a judgement. As long as they don't make claims for how the IR or any other device will sound in someone else's system, they really should be left alone to "have at it". 

 

Note again: this is coming from someone who didn't think the IR made a positive (or negative) difference in his setup. But hey, that was a sighted comparison...so maybe it made a huge positive difference and I couldn't hear it. 

[firedog]

4 hours ago, plissken said:

 

I'm not sure how someone would disagree that Amir just supplying an overwhelming amount of data here didn't just benefit people that are purchasing the Uptone. It does have the downside for Uptone that it may affect their LPS-1 sales. The LPS-1 is most likely gravy. 

Actually, I think Amir's data makes it more likely people will order an LPS-1 or other upscale PS. The device powering the LPS-1 itself is mostly irrelevant. 

[firedog]

4 hours ago, Sal1950 said:

 

The problem arises in the fact that no one from the subjective community will ever participate in real listening tests, so we're left to argue over if some 60 or 70hz noise at -2000 db is audible.  LOL

Not really true. But it's a good myth. And the denial cuts both ways.

Every time someone passes a BT or DBT showing they can hear something supposedly inaudible like low levels of jitter or the difference between hi-res and Redbook, the objectivist camp also claims the test was invalid.  And yes, there have been individuals who have submitted themselves to such tests and passed. 

 

Of course there was also the famous "million dollar" test (the great Randi) to disprove the idea that a high end ethenet cable could make a positive difference. But after the test "proved" that it couldn't, the test conclusions were  shown to be flawed. So it proved nothing, either way.

 

It's convenient to wave around the flag "no one will ever participate", but not so easy to setup a well designed test, especially if you want to test enough individuals of various types to actually prove something. 

[firedog]

1 hour ago, Sal1950 said:

You know that is for the most part total baloney your trying to pass off.

esldude posted numerous tests that people ran away from participating in.

If people were passing these DB tests on magic dust products like you claim, they'd be posted from one end of this site to the other, jumping out of the pages on Stereophile-TAS, proof of purpose demo's being run at all the big Hi Fi shows, etc etc etc.

There have been some well done ones that were passed, like the one Amir took and posted in reply to you on ASR,

http://audiosciencereview.com/forum/index.php?threads/uptone-iso-regen-review-and-measurements.1829/page-17#post-46864

AND the tests your referring to are ones of things that ARE measurable,  not dracula's coffin of dirt grounding schemes.

What particular myth would you like to bust?  The guys might  be able to whip up a opportunity for you.

 

 

Sal, can you please stop just trying to score points? You are arguing and being insulting about claims I didn't make. 

You made a broad generalization about "subjectivists" being in denial.

 

My response was that "objectivists" also exhibit this trait, and I gave  a few examples: Look around the net where it is constantly claimed that "science proves" that listeners can't differentiate between hi-res and redbook, or between moderate and low levels of jitter. 

Yet, it has been shown that trained listeners can do both in properly conducted tests. So you and Amir are only confirming what I said, not the opposite. I'm not sure how we disagree here.

 

In spite of this, when  tests like this come up, many  "objectivists" will repeatedly claim that "the test was flawed", because the results go against what they "know" must be true. So in that they are no different than  the people you claim are in denial.  Just surf a bit and you will find examples. 

 

Beyond that, I posted at ASR a request to Amir that he stop cherry picking my posts here, distorting what I said, and then using them to score debate points. Quite disrespectful, and even obnoxious behavior, IMO. 

 

[firedog]

Just now, Sal1950 said:

Stop squirming firedog, I won't ask you to take any listening tests.

You put up a couple of instances of passed listening tests and use them in general support of the "trust your ears, everything matters" mantra. I can pass a test of my loudness button being on or off too. LOL  The subjective audio world, from both marketers, clients, and media a is full of ridiculous claims of audibility that no one will step up to the plate to support. These things are discussed on this site every day but again no one will provide evidence.

Your straw dog of examples brought nothing of value to the discussion.

Wow, Sal, have you lost your reading glasses?

I've said nothing of the sort. You keep projecting onto my posts things I haven't written. Even when I agree with you and Amir you somehow find it objectionable. Please find one post of mine among the thousands here where I claim that one should only "trust their ears", and ignore everything else. 

If you want to debate me, please debate what I post, not some imagined version that exists only in your head.

[firedog]

1 hour ago, esldude said:

Copy of a response to you over on ASR:

 

Could you specify or link to those if possible. I don't know of any on jitter.

On hires vs redbook there are always, in the ones I know, caveats. And that is okay, just wish to know which ones you speak of.

There is one where hires downsampled was detected vs hires, but not native redbook vs hires. Seems likely the downsampling is what was audible. There are the MQA related results, but they also include poor dither practice and unusually steep filtering, and point to a very weak effect. There was another where DSD was detected vs 48 khz, but not detected vs 44 khz which seems as if something odd must have been going on.

I don't know of any showing jitter being detectable except at very high levels. Would be most interested in those results.

 

Don't have the time to look now. I'm going on vacation and am running out of time to do everything I need to before I leave.

AFAIR, Harman did some tests showing that trained listeners could detect such things. I might have it confused with something else. 

But as noted as ASR, Amir himself says he passed such a test with hi-res.

 

Years ago there was a "jitter test" published online with a file given various levels and also no added jitter. There were people who did testing with these files and succeeded in identifying quite low levels of jitter. As I recall,  I could fairly reliably differentiate the file with 30 ns of added jitter, from the file with 100.  The file with 100 wasn't hard to pick out. Below that it got difficult to hear for me.  Oft quoted by people claiming to represent "science"  is a figure that jitter below 50ns or  250ns is inaudible.

[firedog]

One  problem I see at ASR is that they confuse "taking a measurement" with arriving at a  scientifically proven conclusion. They don't seem to understand the difference. 

 

Another is that many  posters there seem to revel in being nasty jerks. I tried to point out that there is a difference between measuring something - what the posters there insist, for some reason, is the same as a scientific proof - and coming to a scientifically valid conclusion. 

 

For that effort I was repeatedly insulted there, and won't go back.

 

I won't be surprised if this post also gets me some defensive and nasty replies from some of the people that post there....