Science Thread: Review of Audioquest Jitterbug and Uptone Regen USB Conditioners

dallasjustice

Member Sponsor
Apr 12, 2011
2,090
0
0
Dallas, Texas
#81
To those who criticize Amir, I say Amir isn't making claims. Uptone and AQ make plenty of claims. It's totally absurd to shift the burden of proof onto Amir. Amir is simply testing the manufacturer claims made for these two USB hubs. Amir isn't proposing a new theory or offering any product for sale.

Also, Amir's measurements fit nicely with other respected engineers in the field. JA's measurements are very similar to Amir's work. Amir has also previously shown Miller's AQ eye pattern which shows the jitterbug slows the rise time a little. So, it seems like a picture is coming to together.

I know this is a science thread and Amir won't step into the subjective impressions others claim. But I will. I strongly believe most, if not all, who claim an improvement with these USB hubs using decent USB DACs are being subconsciously influenced. These effects are real and play on all humans. All of us are subject to subconscious influence, placebo effect or otherwise. These effects have been carefully studied for decades in the field of psychology. I believe anyone should read the below book by professor nisbett.
Mindware: Tools for Smart Thinking https://www.amazon.com/dp/0374112673/ref=cm_sw_r_awd_6FXcwb4RDP4E6

This book details science's work to date to investigate common thinking errors to which all of humans are susceptible.

Michael.
 

jkeny

Member Sponsor
Feb 10, 2012
3,427
0
0
Ireland
#82
Imagine if Amir had measured a substantial improvement. Then his post would have been paraded about as indisputable proof of the efficacy of these gadgets, no questions asked.
Are JohnW's measurements being paraded as indisputable proof of the effectiveness of these devices? No!
I for one, am waiting for JohnW's measurements on his DAC's analogue outputs when he has done them. Until then what we have from him is that it is effective at greatly attenuating certain noise & jitter on the input USB signal. Finding evidence of it's effect on the analogue out will be challenging, I believe.
 

jkeny

Member Sponsor
Feb 10, 2012
3,427
0
0
Ireland
#84
Do you mean the colourful USB eye pattern? If so, then yes.
Are you denying that JohnW's USB eye pattern plots show the Regen is effective at greatly attenuating the low frequency noise riding on the USB signal & is also improving the USB signal jitter (in his particular setup)?

Is this being paraded as indisputable proof that the DAC's analogue output is different?
 
Nov 27, 2013
217
0
16
#86
Are JohnW's measurements being paraded as indisputable proof of the effectiveness of these devices? No!
I for one, am waiting for JohnW's measurements on his DAC's analogue outputs when he has done them. Until then what we have from him is that it is effective at greatly attenuating certain noise & jitter on the input USB signal. Finding evidence of it's effect on the analogue out will be challenging, I believe.
It is possible. However, if one provides evidence even with dynamic signal, instead of a static one, would you believe it?
I think that many people disregard carefully taken measurements no matter what.
 

jkeny

Member Sponsor
Feb 10, 2012
3,427
0
0
Ireland
#87
It is possible. However, if one provides evidence even with dynamic signal, instead of a static one, would you believe it?
I have always been of the opinion that the effect on the analogue signal will probably be discovered to be in the area of noise modulation at a low level. Any changes in this noise will need to be shown by using a dynamic signal as input & this is why I believe it will be a challenge to measure this - the low level & the dynamic fluctuations both conspire to make it a difficult measurement
I think that many people disregard carefully taken measurements no matter what.
Yes, many don't understand measurements or they dismiss those that aren't supporting their pre-determined opinion.

I had broached this before with Amir in a previous thread about Jitterbug/Regen testing measurements http://www.whatsbestforum.com/showt...g-Measurements&p=334420&viewfull=1#post334420

Originally Posted by jkeny
I'm wondering too how you would go about testing the Regen for which the mode of operation is stated to be dependent on the signal integrity of the USB signal? What would you intend to test for & what test signals used?
It is (relatively) easy to simulate that with extra long cables. Or less than optimal USB host implementations.

As others have asked - have you been able to measure the 8KHz noise spike associated with the USB packet protocol - not that this has anything to do with the Regen's operation but it would show a measurement capability which would lend credence to other USB measurements of a similar nature & level - actually, I believe the effect of bad SI is in producing a modulating noise floor which I think the Regen may well improve (lessen the modulation) - measuring this will be quiet a challenge so I'm interested in your approach.
For this analysis I will try to instrument the USB bus. But for my prior work, all I care about is what comes out of my DAC and USB adapter, and there, as I mentioned, down to -130 dbFS or so, I don't see such noises. Has anyone else measured otherwise on a high performance DAC? I am not saying it doesn't exist but I have only seen USB dependent noise in cheap sync DACs.

Again, I think that this the interesting area that needed investigating in the now closed Measurements thread - what measurements & how they would be approached is the far more interesting (to me, anyway) topic than simply measurements Vs audibility.
I don't know how to resolve that puzzle, nor will I try. I just like to characterize what the devices do or don't from design perspective. There seem to be different approaches with one using a USB hub and the other, analog filtering. There will be differences there of value for some people to know.

I feel a lot of the analysis of measurements is just glossed over because it takes thought, experience & ingenuity to do "proper" measurements (I'm not getting at you, Amir, just stating my thinking about the whole area of audio measurements when it comes to digital audio)
Just want to be clear that while I will be also testing using typical audio measurements, my goal is a circuit analysis at USB level. For sure you can't try to correlate those directly with audibility because what the DAC does with what comes out or doesn't come out of these devices is an unknown.
So it would seem that a bad USB signal is now not being considered (talk of rogue USB ports/computers instead)
It would seem that the 8KHz noise spike I mentioned was completely forgotten about & overlooked in the analysis of the plots - to me they are very evident in any of the Regen plots. To categorise them as jitter spurs is very mistaken, IMO
Some of my comments do seem to have been prescient!
 

Superdad

New Member
Apr 22, 2015
25
0
0
#88
Well here is some hard data for you:
a) The $175 REGEN is sold with a 30-day, money-back, satisfaction guarantee;
b) To-date we have sold 1,723 units;
c) Exactly 8 people have asked for a received a refund on their purchase. That is less than 0.5%.

There is no data there that is relevant to this discussion.

Tim
Why is that data not relevant? Mansr said "might be a fairly strong selection bias in favour of positive reports." I agree in that sure, just like when you look at reviews of products on Amazon, what do you see? Only the people who either love a product or hate a product--those in the generally satisfied middle don't bother to post. Same with restaurants and hotels on Yelp, etc.

But I present hard data (can produce my e-commerce site screen shots if needed) that shows a ton sold and a minuscule asking to return. Money talks, and we know what walks…

By the way, aside from some very mild self-promotion on my own web site's product page--most of which consists of just explaining the mechanism of how the product works and what it is intended to do--there has been close to zero "hype" by us about the product. Zero advertising, zero e-mail marketing, nothing. The REGEN sells entirely by word-of-keyboard, and the for fun push-pin map one of my sons keeps on where I ship to has us up to 45 countries at last count.

If you want to measure jitter, try measuring it at the clock-input pin of the DAC chip. Those J-test output jitter measurements are a joke.

Alex C.
UpTone Audio LLC
 
Apr 3, 2010
16,022
0
0
Seattle, WA
#89
If you want to measure jitter, try measuring it at the clock-input pin of the DAC chip. Those J-test output jitter measurements are a joke.
This is incorrect. The DAC has a PLL which filters jitter. If you measure the jitter prior to that at the clock input, you will be misrepresenting how much jitter will appear in the output of the DAC. It is like saying let's ignore a car has suspension and then see how good it rides. The car has suspension and you can't judge its ride without it. Same with the DAC.

Importantly, what we hear is the analog signal of the DAC. We don't listen to the clock. We have hard enough time convincing people to pay attention to the measurements of the DAC waveform. How would it then be a better approach to then try to get them to understand correlation between some digital clock source and audio???

So no, we don't deviate from industry standards such as measuring the analog output of the DAC. It is the right thing to do.

Now if you are a designer, sure, you instrument the DAC clock jitter. But we are not. We are customers and care about what comes out of our DAC.
 
Aug 6, 2015
30
0
0
Central FL USA
#90
To those who criticize Amir, I say Amir isn't making claims. Uptone and AQ make plenty of claims. It's totally absurd to shift the burden of proof onto Amir. Amir is simply testing the manufacturer claims made for these two USB hubs. Amir isn't proposing a new theory or offering any product for sale.

Also, Amir's measurements fit nicely with other respected engineers in the field. JA's measurements are very similar to Amir's work. Amir has also previously shown Miller's AQ eye pattern which shows the jitterbug slows the rise time a little. So, it seems like a picture is coming to together.

I know this is a science thread and Amir won't step into the subjective impressions others claim. But I will. I strongly believe most, if not all, who claim an improvement with these USB hubs using decent USB DACs are being subconsciously influenced. These effects are real and play on all humans. All of us are subject to subconscious influence, placebo effect or otherwise. These effects have been carefully studied for decades in the field of psychology. I believe anyone should read the below book by professor nisbett.
Mindware: Tools for Smart Thinking https://www.amazon.com/dp/0374112673/ref=cm_sw_r_awd_6FXcwb4RDP4E6

This book details science's work to date to investigate common thinking errors to which all of humans are susceptible.

Michael.
Thank you Michael for the words of wisdom! Sad that so many minds are closed to reality and dismiss any evidence that what they think they hear is in error.
 

marty

Active Member
Apr 20, 2010
1,108
4
38
Far Hills, NJ
#91
This is a wonderful thread and dialog. I found it valuable because I know nothing about digital technology, so I learned a considerable amount. Personally, I welcomed Amir's observations which I thought were accompanied by quite modest comments regarding the limitations and interpretation of his data.

I would add that the following comment left me gasping: "I would not have published your measurements, as it is too easy to interpret them as denigrating a well regarded product. Had they been positive, then I would have been inclined to publish the results..." One may certainly respect a scientist's decision as to when to publish, but as an Associate Editor of the most prestigious journal in my field, I think I can safely say with certainty such a publication policy by any modern medical or scientific journal would never see the light of day. Furthermore, while most high level medical journals consider material for publication that contains (to paraphrase from my own journal) "clinical and laboratory hypothesis-based research with statistically valid results that clearly advance the field" there is also ample opportunity in many journals to report observations and preliminary data that can serve as effective communication and stimulation to others in the field which might stimulate further work on a particular topic. Such observations are widely considered important in fostering collegiality among scientists. I view Amir's OP as serving a similar purpose.

Finally, I am curious however about his comment in the OP about putting a rock on his DAC to make it sound better. What kind of rock exactly? I hope you'll share that with us when you find it as I've tried a few and haven't found one yet that works. :)
 
Jul 1, 2010
8,713
0
0
#92
Sorry, Amir, but it appears that the Science of Audio forum has done nothing to keep the same people from weighing in with the same denials and arguments with the same complete lack of substance. I sometimes wonder why I still click this bookmark.

Tim
 
Apr 3, 2010
16,022
0
0
Seattle, WA
#94
Sorry, Amir, but it appears that the Science of Audio forum has done nothing to keep the same people from weighing in with the same denials and arguments with the same complete lack of substance. I sometimes wonder why I still click this bookmark.

Tim
It is our bad for not being more strict.

Folks, this is the science forum. It is a little pocket we have set aside to discuss audio based on objective and acceptable metrics to audio science and engineering. Subjectivists are welcome and encouraged to participate. But please don't try to argue against data with pure opinion and listening tests not compliant with industry standards. The rest of the forum is wide open to all of those discussions.
 

rbbert

Active Member
Dec 12, 2010
3,423
0
36
Reno, NV
#95
Thank you Michael for the words of wisdom! Sad that so many minds are closed to reality and dismiss any evidence that what they think they hear is in error.
I don't think that is necessarily what is occurring here. I'm not trying to defend subjective listening tests, but for "measurements" to be useful in assessing anything one has to be measuring the "right" things, one has to be measuring them in a repeatable and meaningful manner and one has to be able to interpret the measurements correctly. Many posters are pointing out that in this case none of those requirements have been met. And I'll repeat again, there is nothing "scientific" about this forum so far; no scientists have participated and no accepted scientific method is being used. If you want to be honest call it the "measurement forum", but also be honest about the breadth and depth of ignorance present.
 
Nov 27, 2013
217
0
16
#96
It is our bad for not being more strict.

Folks, this is the science forum. It is a little pocket we have set aside to discuss audio based on objective and acceptable metrics to audio science and engineering. Subjectivists are welcome and encouraged to participate. But please don't try to argue against data with pure opinion and listening tests not compliant with industry standards. The rest of the forum is wide open to all of those discussions.
OT

Amirm, you have PM

/OT
 
Jul 1, 2010
8,713
0
0
#97
I don't think that is necessarily what is occurring here. I'm not trying to defend subjective listening tests, but for "measurements" to be useful in assessing anything one has to be measuring the "right" things, one has to be measuring them in a repeatable and meaningful manner and one has to be able to interpret the measurements correctly. Many posters are pointing out that in this case none of those requirements have been met. And I'll repeat again, there is nothing "scientific" about this forum so far; no scientists have participated and no accepted scientific method is being used. If you want to be honest call it the "measurement forum", but also be honest about the breadth and depth of ignorance present.

Nice.

Tim
 
Apr 3, 2010
16,022
0
0
Seattle, WA
#98
I did not find your work an adequate scientific study for two reasons:
I didn't set out to do a scientific study so I am good then :). As I said, I was curious about performance of these products. Based on what I know these products should make only small difference in measurements but no audible difference in high performance audio products. The praise demonstrated for them though said otherwise. I figured there was some possibility I did not understand the technology and personal investigation was in order and that is what I did.

As you know, I don't get paid for this work, this is not my job, I am not writing a thesis, etc. This is a forum where most of the we exchange opinions. My hope from start has been to inject from time to time, more concrete data to advance our collective knowledge. No standard has been set for those other than being a step above pure opinion exchange.

1. The "mission" is not broken down into separate questions, namely: (a) does the device (e.g. REGEN) improve the USB signal integrity? and (b) does this affect/improve the analog output of the DAC?
Why is this the mission? Who among us is having USB integrity issues for which they bought this product? I am not. Are you? In that regard, (a) is not of any interest whatsoever.

(b) on the hand, is everything and that is where I focused my testing. If (b) shows no improvement I don't care what you did or did not do on the USB bus. As I keep saying, I don't listen to USB bus nor does anyone else.

Once we dismiss (a) as irrelevant mission, then we are this requirement has been met I assume.

(a) For me a scientific study would have required lab quality USB measurement equipment used. Measurements should have been performed to characterize the USB signal quality of a variety of computer sources, involving a variety of computer hardware, computer software and USB cables.
Again, irrelevant. This is not a function that is desired by any user of a DAC. All of our DACs function reliably thank you very much.

So this one doesn't count in our collective application.

(b) For me a scientific study would require a wide range of USB DACs, including those at a wide range of price points, together with comparisons of input waveforms (with and without the REGEN) and output waveforms.
Why? The science says if you have a high fidelity DAC with its own power and asynchronous and isolated USB input, none of these devices are useful or do anything at all. USB is being used simply to carry data much like it would to a printer. Now we could leave it at that and be done with it. Certainly vast majority of professional engineers in AES and such would have laughed at any investigation further than that.

But being more open minded, I chose to spend most of my Sunday doing this testing, spending my own money and that of Michael to confirm what we already know to be true. That is the outcome we got. The science and engineering was proven right. And surprisingly so as these devices instead of making things better, seemingly made things a bit worse.

I was originally going to test with my other DAC but once I got these results, I had little motivation to rip it out of my equipment rack to test. I gave the product the best chance it had for performing: using a low-cost self-powered DAC. If it can't improve that, do you really think a $10,000 would benefit differently?

All of this said, i do have another USB DAC on order. It is the iFi DAC2 or whatever it is called. Assuming Michael doesn't need his Regen back, I may test with that also. But I am telling right now I lack the motivation given the data I already have. I am no longer "curious."

If you all want to collect a bunch of DACs and send them to me, then I might reconsider as I would enjoy testing those DACs on their own. So decide how much you care and not just create work for me with no hope of different outcome :).

In addition the tests should have been formulated to avoid ambiguities, such as the ambiguity of the Jtest signals at 8 and 16 kHz being jitter vs. harmonics. There are enough variables involved that a range of test conditions are needed before it is possible to evaluate any cause-effect relationships. There was no experimental design or possibility of statistical analysis with the small sample of gear tested.
Sorry, there is no issue there. It doesn't matter why those spikes are there. What matters is that they were not there without REGEN. It put them in there and why it did, is something for them to figure out. At best, I expect these products to do no harm. But they did a bit of harm. There are no variables here. There were three possibilities: improvement ,nothing, degradation. We got the latter. Period.

This may be a matter of personal style, but given the results you obtained with what is at best a preliminary investigation, I would not have published your measurements, as it is too easy to interpret them as denigrating a well regarded product. Had they been positive, then I would have been inclined to publish the results as preliminary measurements, but not make any attempt to justify that one is doing any kind of science, or even engineering. What you did strikes me as more appropriately a case of a technician doing preliminary testing of equipment to scope out if it is working, prior to any serious work being done.
You can call the data what you like. I measured what I thought was necessary to establish whether these products do anything meaningful to improve audio performance. They did not. I shared the results with membership along with my personal conclusion. I gave credit and acknowledged that there are many positive subjective reports. I provided both facts and let people decide what it means. I also said there can be human errors in my work and that I welcome manufacturers responding with their own contradictory results. I don't know how much more fair I could be given the outcome.
 

rbbert

Active Member
Dec 12, 2010
3,423
0
36
Reno, NV
#99
And as far as I know completely true. Do you have any evidence to the contrary? Any PhD's in chemistry or physics posting here? Or even PhD's in psychology?
 

rbbert

Active Member
Dec 12, 2010
3,423
0
36
Reno, NV
...You can call the data what you like. I measured what I thought was necessary to establish whether these products do anything meaningful to improve audio performance. They did not. I shared the results with membership along with my personal conclusion. I gave credit and acknowledged that there are many positive subjective reports. I provided both facts and let people decide what it means. I also said there can be human errors in my work and that I welcome manufacturers responding with their own contradictory results. I don't know how much more fair I could be given the outcome.
I'm not sure what the "outcome" is here; perhaps you could state it again. As far as I can tell, you made some measurements; no meaningful correlations were made to their relation to potential audibility (or lack of). Nothing "scientific" about that, any more than someone else's subjective impression that there was (or wasn't) any audible change or improvement with the DUT's. It doesn't become science until you have used your data or observations to formulate a hypothesis, devise and perform an experiment to test that hypothesis, and then analyze those results. In today's world, that must also be followed by peer review and reproducibility, because while you may think that your experimental design and analysis of results (conclusion) is appropriate, your scientific peers may disagree.
 

About us

  • Founded in 2010 What's Best Forum invites intelligent and courteous people of all interests and backgrounds to describe and discuss the best of everything. From beginners to life-long hobbyists to industry professionals we enjoy learning about new things and meeting new people and participating in spirited debates.

Quick Navigation

User Menu

Steve Williams
Site Founder | Site Owner | Administrator
Ron Resnick
Site Co-Owner | Administrator
Julian (The Fixer)
Website Build | Marketing Managersing