Discussion of the UpTone / J.Swenson EtherREGEN 'white paper'

John Swenson spent his 30+ year career as a senior engineer for a major silicone chip firm (LSI Logic/Avago/Broadcom), and his speciality was designing the power networks inside massive ASICs (the sort of large specialty chips inside your computers, phones, and networking gear).  He has also spent considerable time thinking about and designing digital audio-related circuits.

For too long there has been great skepticism surrounding the relevance--or even existence of--small perturbations of jitter and noise on various digital audio interfaces.  Yet of course there are already thousands of people who, in their own music systems, with their own ears, have heard the impact (sometimes subtle, sometimes large) that result from various USB, Ethernet, and power supply changes.

Some will say you are all delusional, and that you are imagining what you hear.  They are wrong.  The effects are real, but the mechanisms behind them can be a little difficult to understand. So in an effort to educate and shed some light on a few of the factors involved, we are presenting what we hope is an easy to digest introduction to the subjects of ground-plane noise, threshold jitter, and leakage currents—and why they matter and have an effect on various receiver chips and the DAC itself.

This paper does not yet present measurement data and graphs. John has—with a custom low-noise input board feeding his Wavecrest timing analyzer—already captured analyzed data showing phase-noise variations at the DAC clock input pin with and without an EtherREGEN in the chain (which was EtherREGEN>ultraRendu>USB DAC). He is working on the graph axis and legends and he will present these and discuss the measurement system in the coming weeks.

Yet even with published measurement evidence, those who don’t believe any of this stuff matters will then simply claim that what we are measuring can’t possibly be heard. Someday the things we are discovering and dealing with now will be accepted and seen as commonplace—just as germs and jitter were discovered. A quote that struck me in a TED Talk by an open-minded scientist sticks with me: “Dogmatic assumption inhibits inquiry.”

As most of you here know, I, Alex Crespi, am not an engineer, nor do I pretend to be. And while I spent a lot of time editing and organizing this paper--to the point where I have mentally internalized most of its concepts--I am clearly not going to be the right person to answer probing questions or challenges to it.  John Swenson is the man for that. Yet while I am sure he will regularly chime in with cogent answers, please recognize that his time is limited and valuable. He is actively working on additional groundbreaking products--which will eventually come to market under both the UpTone and Sonore brands (different products).

Enjoy the paper, and also the very fine and thoughtful review of the EtherREGEN published just today on the homepage of Audiophile Style. 

------------------

Here is a LINK to the PDF of our 5-page paper for offline reading.  And below are the JPG images of the individual pages:

Congrats to John and Alex for publishing this White Paper. I know that it was a lot of hard work and effort. Its very informative and reads well, so well done. 👍

On 3/10/2020 at 9:17 AM, Superdad said:

John has—with a custom low-noise input board feeding his Wavecrest timing analyzer—already captured analyzed data showing phase-noise variations at the DAC clock input pin with and without an EtherREGEN in the chain (which was EtherREGEN>ultraRendu>USB DAC). He is working on the graph axis and legends and he will present these and discuss the measurement system in the coming weeks.

Yet even with published measurement evidence, those who don’t believe any of this stuff matters will then simply claim that what we are measuring can’t possibly be heard. 

15 hours ago, simon_pepper said:

Thank you for an interesting read and explanation of the issues, causes and methods used to address. However the paper would benefit from some measurements, before/after analysis of both the jitter handling and AC ground plane noise.

I understand these measurements are hard to setup and ensure are correctly showing the problem and solution, but the positioning outlined in the text would greatly gain further weight from these.

Hi Simon:

As indicated, John has made measurements (and is making some more including ground-plane noise), and we will be presenting these. Yet we know for certain that even once we do, our most savage critics will not be satisfied.  They are too entrenched in their "flat earth" world view. 

21 hours ago, Superdad said:

 

As indicated, John has made measurements (and is making some more including ground-plane noise), and we will be presenting these. Yet we know for certain that even once we do, our most savage critics will not be satisfied.  They are too entrenched in their "flat earth" world view. 

I see this a little differently.  It is not really about the very small number of "savage critics", more about the far more numerous "ordinary folk" in the middle.  If I wanted to, I know enough about audio streaming to put together a very convincing argument as to why something like an audiophile network switch cannot possibly make any difference to sound quality.  If I wanted to, I could easily put together an argument as to how expectation bias / confirmation bias can explain all the positive listening reports.  But I am not going to do that, because I am one of the very many "ordinary folk" in the middle.  Although I understand what might be called the "objectivist" arguments, I am also open minded, interested in the new theories re phase noise etc, and interested in trying new technology such as the EtherRegen.  Indeed, I have bought one, and I have to say I am very happy with it, I think it has improved the sound quality of my system, I am enjoying the music more now, less distracted by "digital hash".  But, because I understand the the kind of arguments made by ASR and others, and because I understand expectation bias, I do sometimes wonder myself if this stuff is real, am I myself imagining the sound quality improvements?  To be honest, I do not think I am imagining it, but I do still wonder sometimes, it is not impossible if you consider that expectation bias is a powerful thing.  So, for the very many reasonable, considered, ordinary "folk in the middle", some measurements would be most welcome, it would be so nice to have some firm data to back up what we are hearing.  As things stand, the "savage critics" do have their own measurements, it would be good for us ordinary folk to have some too.  As for satisfying the savage critics themselves, maybe this is impossible, but at the end of the day they are not your customer base anyway.

Anyway, it is good news that you have some measurements, and I understand why you may need more time to finalise before publishing.  It's all good!

6 hours ago, Superdad said:

 

 

 

Hi Simon:

As indicated, John has made measurements (and is making some more including ground-plane noise), and we will be presenting these. Yet we know for certain that even once we do, our most savage critics will not be satisfied.  They are do entrenched in their "flat earth" world view. 

Alex,

I look forward to seeing John's measurements in due course.

There are definitely critics with strongly entrenched views who don't believe that the ER is, or can be, effective. We've had that problem in the past in audio on lots of things including differences in harmonic distortion in amplifiers when the only measurement conducted was THD and nobody measured the harmonic spectrum of the distortion, and with early CD players with identical distortion measurements in the days before anyone started measuring jitter. It's impossible to resolve disputes related to the absence of measurements showing a difference if the characteristic that differs isn't being measured, and reviewers may not measure the right things if they don't know what to measure and how to measure it.  We do need to see John's measurements simply in order to know what measurements are important and how they should be made if we're to know what measurements we should rely on as an indication of the ER's performance.

When it comes to listening tests/reports, all we have are the impressions of a number of people, including the author of the ASR review. While most of those are positive there are some negative impressions amongst them. All of those impressions are based on sighted tests, even the impression of no difference reported in the ASR review. What makes a test reliable is the consistency of results when different people conduct the same test using the same procedure. While I doubt that the differences in procedure associated with the comparisons most of us have made between the ER and other switches are significant, the fact that some people report hearing a difference and some report hearing no difference means that we don't have the consistency in results to make those listening reports reliable at a scientific level. What would be required for that is two or more professionally conducted blind or double blind tests conducted with the same or comparable test protocols and with a reasonable number of subjects in each test, which produce similar results. Personally I find the preponderance and general consistency of the positive listening reports (most especially my own) compelling even though they fall short of the level required for the scientific certainty demanded by some.

Since I have doubts that we're likely to see such listening tests anytime soon (they're expensive and time consuming to run), I think that it's important to see John's measurements along with details of his testing methods so that others can repeat his procedures and tell us whether his results are replicable.

Even then, I suspect that there will be some critics who would dismiss the results out of hand but then I do know of some people who still think that Edison cylinders and 78s have their charms, especially when played on an original acoustic machines  🙂

In the meantime I'm convinced by the sound I'm hearing with the ER in my system and enjoying it greatly. I should also add that I found John's paper both interesting and informative and I'm glad to see it published at last.

David

4 minutes ago, David A said:

--David

Brilliant David. I agree with everything you just said.

John is working on both documenting his sensitive phase-noise at the DAC clock-pin test methods and on producing ground-plane noise measurements.  He has been doing measurements with this project for some time, but putting them into a form with scales and legends that are relatable for others is only now being done.  

UpTone and Sonore are pretty demanding of his time on development projects moving forward, and, as there are no direct profits from days/weeks of painstaking measurements, his time on such stays limited. It's not as if we have a whole staff of brilliant engineers. Just the one.  

During over 50 years in this hobby, I have learned that Everything Matters. I wish that it didn't, but I have repeatedly found that it does (though admittedly not always for the better--pick your fights). Rajiv's masterful review is testament to this notion.

The psychological term "expectation bias" gets thrown around a lot around these parts, but remember that it cuts both ways. I'm referring to those who reflexively  "expect" that nothing short of "upgrading" a source component, an amplifier or a loudspeaker can possibly make a difference in how we subjectively experience and interact with our favorite music in our homes. Even to the extent of using limited or irrelevant measurements to support this particular form of bias (yet not even trying something interesting out there despite ubiquitous offers of free trials),

Expecting no audible changes when swapping or adding any particular piece of your system is likewise a form of bias. Dismissing any thoughtfully engineered product with good reviews / word of mouth out of hand, just seems dumb to me. It suggests that these folks are not really serious about realizing the greatest degree  emotion and joy when experiencing their favorite music. 

Lately, I have been listening to AudioStream podcasts on my daily hikes, most recently interviews with the likes of digital luminaries such as Ted Smith (PS Audio), Rob Watts (Chord), Gordon Rankin (Wavelength), Vince Galbo (MSB) and Bob Stuart (Meridian/MQA), And certainly put John's White Paper in the mix as well. All are pursuing perfection in technically very different ways, but pay close attention and you begin to realize that they are all seeking to deliver much the same thing in the end.

Over very long careers, they have developed deep experiential knowledge about why Everything Matters. 

• Why noise can ride on the signal impacting not the bits, but how--and more importantly when--the shift from zero to one occurs.
   
• Why humans can detect the absence of transient and harmonic content that is well above the limits of our hearing in the frequency domain. 

Few of my friends are audiophiles, but to a one they love music. And they all love to enjoy their favorite music at my house.

Yeah at first they might roll their eyes at the array of doodads large and small spilling out onto the floor (hey, I'm a bachelor), but when they're in the sweet spot and you hand them the iPad controlling Roon, you know that they're gonna be there for a while, all smiles. And why, after they depart, I might spend another hour or so continuing the concert.

Finally,in a last ditch attempt to get this rant at least semi back on topic, in a few short months there have been multiple dozens of folks reporting remarkably consistent impressions of what the ER brings to the streaming party. With 1200 sold (I was in the first batch) and rarely does an Uptone Audio appear for sale on Audiogon. Kinda tells you something.

If it's an hallucination, it's a mass hallucination!!!

Andy

2 hours ago, so-no-mah said:

During over 50 years in this hobby, I have learned that Everything Matters. I wish that it didn't, but I have repeatedly found that it does (though admittedly not always for the better--pick your fights). Rajiv's masterful review is testament to this notion.

 

The psychological term "expectation bias" gets thrown around a lot around these parts, but remember that it cuts both ways. I'm referring to those who reflexively  "expect" that nothing short of "upgrading" a source component, an amplifier or a loudspeaker can possibly make a difference in how we subjectively experience and interact with our favorite music in our homes. Even to the extent of using limited or irrelevant measurements to support this particular form of bias (yet not even trying something interesting out there despite ubiquitous offers of free trials),

 

Expecting no audible changes when swapping or adding any particular piece of your system is likewise a form of bias. Dismissing any thoughtfully engineered product with good reviews / word of mouth out of hand, just seems dumb to me. It suggests that these folks are not really serious about realizing the greatest degree  emotion and joy when experiencing their favorite music. 

 

Lately, I have been listening to AudioStream podcasts on my daily hikes, most recently interviews with the likes of digital luminaries such as Ted Smith (PS Audio), Rob Watts (Chord), Gordon Rankin (Wavelength), Vince Galbo (MSB) and Bob Stuart (Meridian/MQA), And certainly put John's White Paper in the mix as well. All are pursuing perfection in technically very different ways, but pay close attention and you begin to realize that they are all seeking to deliver much the same thing in the end.

 

Over very long careers, they have developed deep experiential knowledge about why Everything Matters. 

• Why noise can ride on the signal impacting not the bits, but how--and more importantly when--the shift from zero to one occurs.
   
• Why humans can detect the absence of transient and harmonic content that is well above the limits of our hearing in the frequency domain. 

Few of my friends are audiophiles, but to a one they love music. And they all love to enjoy their favorite music at my house.

 

Yeah at first they might roll their eyes at the array of doodads large and small spilling out onto the floor (hey, I'm a bachelor), but when they're in the sweet spot and you hand them the iPad controlling Roon, you know that they're gonna be there for a while, all smiles. And why, after they depart, I might spend another hour or so continuing the concert.

 

Finally,in a last ditch attempt to get this rant at least semi back on topic, in a few short months there have been multiple dozens of folks reporting remarkably consistent impressions of what the ER brings to the streaming party. With 1200 sold (I was in the first batch) and rarely does an Uptone Audio appear for sale on Audiogon. Kinda tells you something.

 

If it's an hallucination, it's a mass hallucination!!!

 

Andy

After over 50 years in this hobby also, I have a slightly different take on "expectation bias" cutting both ways.

Expectation bias certainly exists but it applies to all expectations. Certainly those of us who expect to hear a difference from something can make the mistake of hearing a difference that doesn't exist but that certainly doesn't happen all of the time, and sometimes the difference we hear can be quite different to the difference we expected. On the other hand, those who expect not to hear a difference can make the mistake of failing to hear a difference that does exist and I've seen more than a few examples of that. I don't believe those who expect not to hear a difference are immune to the effects of expectation bias. Whatever our expectation, we can all make the mistake of hearing what we expect when the opposite is the case. No one is immune to expectation bias, regardless of the nature of their expectation.

No one spends years in this hobby without being surprised and having their expectations dashed some of the time, regardless of their expectations and beliefs, and no one spends years in this hobby without getting it wrong some of the time. We're all human and prone to making mistakes because of our beliefs and expectations a few times over the years. I know I've made that kind of mistake more than once and I'm old enough to be able to avoid telling you how many times I've made that mistake by saying I can no longer remember all of my mistakes 🙂

In science observations, if confirmed, trump accepted knowledge and belief every time. The big problem is often simply confirming the observations, regardless of their nature. Really good equipment designers are also probably good observers who take the time to confirm their observations, come up with good explanations for the observations they did confirm, and then make use of that understanding in the gear they design. That's the practical application of the scientific method.

And, like you, I'm familiar with the sight of visitors rolling their eyes when they first see my system and then sitting there with a silly grin on their face when they start to listen. That's one of the joys of being an audiophile and one of the nicest signs of observations that don't agree with expectations. Audiophiles should spend more time surprising others, audiophiles or not, rather than disparaging each other with accusations of expectation bias because every one of us is prone to that.

David

9 hours ago, so-no-mah said:

...in a few short months there have been multiple dozens of folks reporting remarkably consistent impressions of what the ER brings to the streaming party. With 1200 sold (I was in the first batch) and rarely does an Uptone Audio appear for sale on Audiogon. Kinda tells you something.

If it's an hallucination, it's a mass hallucination!!!

Oh no, EtherREGEN could be the corona virus of audio 2020!  It’s real, it’s spreading fast, and it should be taken seriously. 

[Apologies if this seems in bad taste. Believe me, due to my wife’s work, Covid-19 is a somber and dire topic around our household this week. As for many around the world. Makes audio seem a bit frivolous at the moment.]

13 hours ago, Confused said:

I am one of the very many "ordinary folk" in the middle.  Although I understand what might be called the "objectivist" arguments, I am also open minded, interested in the new theories re phase noise etc, and interested in trying new technology such as the EtherRegen.  Indeed, I have bought one, and I have to say I am very happy with it, I think it has improved the sound quality of my system, I am enjoying the music more now, less distracted by "digital hash".  But, because I understand the the kind of arguments made by ASR and others, and because I understand expectation bias, I do sometimes wonder myself if this stuff is real, am I myself imagining the sound quality improvements?  To be honest, I do not think I am imagining it, but I do still wonder sometimes, it is not impossible if you consider that expectation bias is a powerful thing. 

I think your "man in the middle" is what is known as a scientist.

Doubt is what drives science and IMO that includes doubt in oneself. People who are certain of everything have faith. You might call them extreme subjectivists or extreme objectivists but they worship at the same altar 😉

Ah, I think I was the first one to mention expectation bias in this thread.  OK - One point to clarify, in the context of my earlier post I was referring to my own expectation bias.  So I buy an EtherRegen.  Why?  Because I think it will improve my system.  I think this because of what I have read about the EtherRegen, the concept and the technology, and by the time I actually have the EtherRegen and install it in my system, I have read quite a few positive reports from those that received one before me.  So on first use, I am both optimistic and hopeful that it will offer an improvement.  Can this influence my own expectation bias?  Maybe, I am certainly not arrogant enough to claim I am 100% immune to it.  As it happens, I was similarly optimistic when trying an AQVOX switch, but ultimately decided it was not doing anything positive for sound quality and so I returned it for a refund.  Based on this, I could claim I have the power to overcome expectation bias, how else to explain that I hear a difference with the EtherRegen but not an AQVOX switch, when in both cases I was expecting an improvement?  The thing is, human psychology is a very complex thing, the argument may seem sound but the reality is far too complex to analyse in this way.  

A lot of waffle here, but the point is exactly as @Audiophile Neuroscience is alluding to in the post above.  I know enough to doubt myself, which is why I would particularly like to see the white paper backed up by measurements.  I am genuinely looking forward to seeing the measurements, I am sure it will be fascinating. 

1 hour ago, Confused said:

the point is exactly as @Audiophile Neuroscience is alluding to in the post above.  I know enough to doubt myself, which is why I would particularly like to see the white paper backed up by measurements.

Personally, I can live with doubt. I am uncertain of many things. The good guys in this equation IMHO are people like Alex C and John S who dare to disagree with "the impossible" and imagine that the world is not flat after all. Of course more evidence is welcome but my problem is I am currently listening to the fruits of their labour and I am enjoying it. If it is an hallucination it is a very good one.🥰

4 hours ago, Superdad said:

As for many around the world. Makes audio seem a bit frivolous at the moment.]

Maybe but for me the music is a welcome distraction of peace form 12 hours a day of corona speculation, mis information and Facebook wisdom of my colleagues.

4 hours ago, firedog said:

 

Am I missing something here? The critics' demand has always been for measurements showing the analog output of the DAC is affected by the various types of noise. Showing that the noise arrives to the DAC isn't going to answer that question. 

Correct, I'm working on DAC output as well. None of this stuff is easy. We are trying to measure a DAC output using ADCs, to do that requires an ADC that has lower clock jitter than the DAC we are trying to measure, this is NOT easy to do! The off the shelf audio analyzers are nowhere near close enough. So as with everything else related to this I have to make my own. It's not easy and it's not cheap.

I have something up and running which still needs a lot of more work to lower ADC jitter and prevent external signals from getting in, but it DOES actually get signal through. Maybe within the next week I'll try it with real audio DACs and see how things are at this point in the development process.

My ultimate goal is to show the whole process from ground plane noise, to DAC clock jitter to DAC analog output.

I WILL get there, but it is going to take a significant amount of time to get there.

John S.

There are some fundamental points here that I'd like raise, but just once, for folks to think about and consider, but I won't really get into deep discussions about these points, as things will stray off topic. But I'm putting them forward just to level set. 

The basic premise is that there are many things in this "universe" and in the experiences of our lives that we know to be TRUE and REAL that reductionist science cannot explain, either with "measurements" or even at a deeper fundamental level. And because reductionist science cannot explain them, they are...ignored. 

1. Gravity: Physicists have no explanation for Gravity. Yes, it can be characterized by Newton's famous equation, F=GMm/r2 that mathematically describes the force of gravity between two masses, but no physicist can explain WHY two masses are attracted to each to eather through a gravitational field. In fact, physicists cannot actually tell you what a field is and how it is mediated through "nothingness". This is an extension of another thing they can't explain, what is known in quantum physics as "the hard problem". 
2. Gravitational Constant: Furthermore, physicists can't explain why the gravitational constant, G, changes over time, and varies.  Guess what? If the gravitational constant is a constant, why does it change over time? 
3. Speed of Light: the speed of light has changed over time as well. If you look at measurements in the historical record, it is documented that the measured speed of slight dropped between 1930 and 1945. If the speed of light is a constant, why does it change over time? 
4. Magnetism: nobody knows how a magnet can move a piece of metal without touching it. And for another thing is, nobody seems to care. When Richard Feynman was asked why by an interviewer: "If you hold two magnets with the same poles close, you'll feel a force pushing them away, and if they have opposite poles, they will snap together. "What I want to know is, what's going on between these two bits of metal?", he could not provide an answer. Because he didn't know more than anyone else does. Some physicicsts use "virtual photons" to explain magnetism, but "virtual photons" don't actually exist, they are simply mathematical fudge factors that theoretical physicists use to try to balance their equations. 
5. Dogs (and many cats) Wait for Their Owners to Come Home: Millions of dog and cat owners have experienced that their dogs, and to a lesser degree cats, are always waiting for them to come home.  It turns out they are waiting for them before the owner actually arrives at home. This is independent of the mode of transport of the owner (on foot or with mechanized tranportation and independant of time of day, schedule, etc.) This phenomenon has been documented using professional video camera crews with time-synchronized cam-corders at the location of the dog and independently and simultaneously at the location of the owner, who is away from home, with statistical signficance. Science has NO answer for this well-known and well-characterized experience.
6. The Sensation of Being Stared At: The sensation of being stared at is an experience that virtually all of us have had, or experienced from staring at someone who is not looking at them (a simple experiment: when you are driving down the freeway, turn and look at a driver in a car alongside you that is not looking at you, and measure how often, when you look at them, they will turn their head and look back at you). A simple experiment for this phenomenon has been conducted in studies world-wide and have been shown to be real with a p-value that provides statistical significance.  As of 2005, there had already been 30,803 trials at multiple labs and locations of subjects being stared at with 16,849 correct observations, producing a p-value of 1X10^-20 (thus, the probability of obtaining these results by chance are significantly less than 1 in ten billion) Ref: J. Conciousness Studies 12: 10-31 (2005).
7. The Collapse of the Wave Function in a Double-Slit Experiment. The collapse of the wave function (interference pattern) of light in a double-slit experiment when viewed by a Conscious Observer is a well-characterized phenomenon. This phenomenon is held even when the conscious observer is viewing remotely by video display; it does NOT occur when the observing is done by a computer. Dean Radin has conducted an outstanding set of experiments producing the same p-values that won the Nobel Prize for the discovery of the Higgs Boson. Physicists and "Science" have no answer for this phenomenon. YouTube Video here: https://youtu.be/nRSBaq3vAeY

The excellent white paper that John has written provides the foundation for understanding how a new class of noise factors e.g. the different jitter classes, the impact of leakage currents and clock phase noise and their interactions impact the reproduction of digital music. Impacts it in a way we hear and know to be true in an audible manner.  I'm confident as we learn more about digital music reproduction, we will discover and characterize more of these factors.

Regardless, though, there are many phenomena that I've described above that are known and documented to be true that quite simply, reductionist science cannot explain. Simple stuff, you know, like Gravity and Magnetism. But we DO know these phenomena to be true. The fact we can't explain or measure them does NOT make them any less true.

The same applies to what we can hear, and the interaction of our audio equipment, our brains, minds, and what we know we can hear when listening to music.