UpTone Audio EtherREGEN (Objective Discussion Only)

[plissken]

2 hours ago, charlesphoto said:

He's laid out the reasons, though much of it is over my head, but better power supply, PHY layer, IGMP snooping etc. So pick up a used one, and compare (the newer the better many report).

 

Simon is making a few claims/points.

 

1. Cisco uses tighter tolerance 25Mhz clocks

2. Enabling IGMP Snooping reduces noise. In  a nutshell IF you have client applications that have registered themselves with a 224.0.0.0-239.255.255.255 address they are registering for a multicast (there are reserved addresses like 224.0.0.5/6 for OSPF DR/BDR etc...). There is a wiki page for this if you are interested.

 

Unconstrained multi-casts can quickly dominate a subnet, especially large subnets (we typically will do /20 for guest wired and wireless). We control multi-casts and I'd rather RACL and VACL on larger subnet than multiple smaller ones.

 

What happens w/o snooping is the multi-casts will get forwarded out all ports (think about a subnet with 8,000 clients). You will have clients that didn't register  for a multi-cast get this traffic and each of these clients will filter / drop.

 

IGMP snooping in a nut shell builds a database of clients and what they are registering for and does the dropping action for the clients. So multi-casts only end up going to registered ports.

 

I certainly recommend the 2960's, I am also a huge fan of the 2390: You get 10GBe SFP+ and you can get the switches for $60. Incredible value.

 

1st off I recommend Cisco because it's simply well built. I've yet to hear them make an audible difference in my setup. I'd rather spend $60 on a used/refurb Enterprise class edge Cisco than $60 on a new Netgear. Although I've no complaints about all the Netgear GS series I've setup. Bullet proof and reliable as I could have ever wanted.

 

2nd off IGMP Snooping just isn't much use on a home network. If you have a Cisco switch just issue 'Sh int interface#here' and 'Sh buffers'. You'll find your switch is WAY underutilized, and 0 errors with proper cabling and well behaved NIC's.

 

I setup IGMP v3 and PIM-Sparse in my environment with systems like Get Well for delivering patient informational videos in addition to guest vlan.

 

[Superdad]

5 hours ago, charlesphoto said:

I do know Uptone broke one of these down in researching the eR. No clue how they measure. Don't care as they improved the sound of my system.

 

2 hours ago, plissken said:

1st off I recommend Cisco because it's simply well built. I've yet to hear them make an audible difference in my setup. I'd rather spend $60 on a used/refurb Enterprise class edge Cisco than $60 on a new Netgear. Although I've no complaints about all the Netgear GS series I've setup. Bullet proof and reliable as I could have ever wanted.

 

Having nothing to do with multicast, buffers, or any protocol stuff, there are some very specific circuit design elements that made the Cisco Catalyst 2960 a "good sounding" switch.

But since the skeptics in this thread are only interested in finding the results in the "poop," I'll not speak further about the ingredients that went into cooking the "soup." 

[plissken]

23 minutes ago, Superdad said:

Having nothing to do with multicast, buffers, or any protocol stuff, there are some very specific circuit design elements that made the Cisco Catalyst 2960 a "good sounding" switch.

But since the skeptics in this thread are only interested in finding the results in the "poop," I'll not speak further about the ingredients that went into cooking the "soup." 

 

I still have my EMU 1212 PCI-e sound card. It has 1/4" TRS balanced I/O. I was able to purposefully construct an XLR cable and you could hear HDD or SSD R/W, mouse movement etc...

 

I guess I could try this with some various switches and see if that is going on. I'll have to get an older computer that can take a Windows 7 install as that is the last OS Creative had 100% support for.

 

EDIT ***I ordered a system from NewEgg***

 

 

 

 

[jabbr]

6 hours ago, Audiophile Neuroscience said:

In essence this is what has been done.People have listened and made observations. Measurements are apparently pending.As pointed out by @Iving the process is a very complicated one from an objective scientific viewpoint.

 

There are some very simple measurements that could be made ...

 

For example:

1) noise on the ground plane of the EtherREGEN vs Cisco vs Trendnet

2) noise on the receiver ground plane e.g. microRendu using EtherREGEN vs Cisco vs Trendnet

 

None of that needs very fancy equipment to measure if one wanted to do the measurement.

[Superdad]

4 minutes ago, jabbr said:

There are some very simple measurements that could be made ...

 

Very much along the lines of what we will publish. Though as you know, it takes careful attention to environmental factors to properly measure very low level noise.  And the perturbations we are looking at do not need to be very large to have the effect on clock threshold jitter that we believe are the root cause of the sonic differences heard.

 

@JohnSwenson's expensive new PhaseStation is showing just how much environmental factors influence low level phase-noise measurement. DC cables, vibration, fields in the air, lighting, even body presence are causing wild wiggles at the levels he is testing. He is building cases and supplies to reduce those distractions. And we are not even talking about your favorite, 1/f noise. 

[Audiophile Neuroscience]

11 minutes ago, jabbr said:

 

There are some very simple measurements that could be made ...

 

For example:

1) noise on the ground plane of the EtherREGEN vs Cisco vs Trendnet

2) noise on the receiver ground plane e.g. microRendu using EtherREGEN vs Cisco vs Trendnet

 

None of that needs very fancy equipment to measure if one wanted to do the measurement.

I agree Jonathon.

 

I would welcome these or any other relevant measurements. The keyword here I guess is relevant. For me, I would be interested in any measurement of the audio signal that shows a difference.  I do agree however with @Iving that the process does not stop there. Indeed I would anticipate that others will argue that the process does not even start there and specifically, the measured change in the audio signal must be at the output of the DAC, anywhere else being irrelevant.

[jabbr]

7 minutes ago, Superdad said:

 

Very much along the lines of what we will publish. Though as you know, it takes careful attention to environmental factors to properly measure very low level noise.  And the perturbations we are looking at do not need to be very large to have the effect on clock threshold jitter that we believe are the root cause of the sonic differences heard.

 

I am assuming that there should be a rather large and easily identifiable difference in the ground plane noise between your device and a generic low cost switch. 

7 minutes ago, Superdad said:

 

@JohnSwenson's expensive new PhaseStation is showing just how much environmental factors influence low level phase-noise measurement. DC cables, vibration, fields in the air, lighting, even body presence are causing wild wiggles at the levels he is testing. He is building cases and supplies to reduce those distractions. And we are not even talking about your favorite, 1/f noise. 

I hope that some easy measurements would convincingly show something, rather than chasing ghosts. 

[plissken]

29 minutes ago, Superdad said:

 

@JohnSwenson's expensive new PhaseStation

 

Alex, when you praise, or denigrate a piece of equipment you seem to toss in something about it's relative price.

 

Just saying....

 

 

[Superdad]

 

12 minutes ago, plissken said:

Alex, when you praise, or denigrate a piece of equipment you seem to toss in something about it's relative price.

 

On 3/31/2020 at 3:14 PM, plissken said:

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

 

Pot, meet kettle. 

 

But seriously Mark, what is it you expect to measure or record with your 15 year old EMU PCI-e sound card? The one with the ground loop picking up mouse/drive interference inside your PC tower case. 

[jabbr]

59 minutes ago, Audiophile Neuroscience said:

I would welcome these or any other relevant measurements. The keyword here I guess is relevant. For me, I would be interested in any measurement of the audio signal that shows a difference.  I do agree however with @Iving that the process does not stop there. Indeed I would anticipate that others will argue that the process does not even start there and specifically, the measured change in the audio signal must be at the output of the DAC, anywhere else being irrelevant.

 

No doubt that measurements at the DAC output are what is ultimately desired however there are things about network/usb phase noise that have been bandied about for about 5 years now without even the most basic measurements.

 

My experience is that when someone is having trouble demonstrating something like this with measurements, it means that whatever measurements have been done don't support the theory.

 

My own experience with audible ground loops is that they are easy to measure. We aren't looking for Bs meson's here  

 

I am eager to be ediucated.

[jabbr]

1 hour ago, Superdad said:

 

@JohnSwenson's expensive new PhaseStation is showing just how much environmental factors influence low level phase-noise measurement. DC cables, vibration, fields in the air, lighting, even body presence are causing wild wiggles at the levels he is testing. He is building cases and supplies to reduce those distractions. And we are not even talking about your favorite, 1/f noise. 

 

The other thing is that really low level changes such as this are much less likely to be audible than large easily measurable differences in phase error from really cheap to reasonably good clock oscillators.

 

For example, I can't hear an easily audible improvement with my Mellanox switch (phase error <60 femtoseconds) compared with an old Brocade switch (VDX 6720 with perhaps 200  femtoseconds phase error) -- and that's end to end, rather than merely the clocks!

[plissken]

50 minutes ago, Superdad said:

But seriously Mark, what is it you expect to measure or record with your 15 year old EMU PCI-e sound card? The one with the ground loop picking up mouse/drive interference inside your PC tower case. 

I want to see if any noise is to be had from a network interface transfer with either the inboard NIC or Intel PCI-e NIC and see if my 8 port 1GBe D-Link or one of my Cisco or Aruba switches make a difference in radiated noise.

[manueljenkin]

2 hours ago, Superdad said:

 

Very much along the lines of what we will publish. Though as you know, it takes careful attention to environmental factors to properly measure very low level noise.  And the perturbations we are looking at do not need to be very large to have the effect on clock threshold jitter that we believe are the root cause of the sonic differences heard.

 

@JohnSwenson's expensive new PhaseStation is showing just how much environmental factors influence low level phase-noise measurement. DC cables, vibration, fields in the air, lighting, even body presence are causing wild wiggles at the levels he is testing. He is building cases and supplies to reduce those distractions. And we are not even talking about your favorite, 1/f noise. 

An electrical/electronics engineer myself, I'm really curious to know about the aberrations present. Any article/post here or on your webpage would be very helpful for me. My wish is to be a part of a team that does instrumentation for medical/science purposes, and knowledge across these domains would be of great interest. I really don't care about correlation to audio, I'm more interested in its utility for ANY realtime high precision application (which can include audio to some extent).

 

I also sent a mail to shunyata research asking guidance in Power supply design. I'm copy pasta-ing the post here, kindly guide me if any of those are relevant to er or is something you're familiar with.

 

" I have seen a few of your videos, and I am interested to know more about the dynamic power consumptuion properties of transistors (inrush etc). I am unsure of where to ask for guidance, everywhere I ask for guidance, a dozen vocal skeptics (with absolutely no industry background) come in and say there is no difference, when in reality I could even tweak my software buffer to make a difference in sound. I am well aware of clock skews, oscillator drifts due to unstable power supply, opamps behaving as antennas to rf noises and have also worked as a physical design engineer for a short while (majority was as validation engineer for RTL). I'd love to learn more in detail, not as much as company confidential stuff, but to some extent of depth. Kindly guide me.

Id also like to know if the power filters can bring in improvements to my led lighting brightness/flicker consistency when used in parallel to the wiring for the lights. At the current moment, I am unable to afford any of your systems, so I'm sorry if that was disappointing. But I would eventually buy one of yours products, when I get necessary funds.

On a side note, I'd also love to know guidance on battery parameters (designing my "transportable" amplifier.") Better description here : https://www.diyaudio.com/forums/power-supplies/351812-types-batteries-transient-steady-properties-aberrations-2.html#post6138552 "

[jabbr]

24 minutes ago, manueljenkin said:

An electrical/electronics engineer myself, I'm really curious to know about the aberrations present. Any article/post here or on your webpage would be very helpful for me. My wish is to be a part of a team that does instrumentation for medical/science purposes, and knowledge across these domains would be of great interest. I really don't care about correlation to audio, I'm more interested in its utility for ANY realtime high precision application (which can include audio to some extent).

 

I also sent a mail to shunyata research asking guidance in Power supply design. I'm copy pasta-ing the post here, kindly guide me if any of those are relevant to er or is something you're familiar with.

 

Did you hear? CERN just bought out the remaining stock of Shunyata, for use with the Large Hadron Collider ...

 

No seriously, EMI labs etc, and many which measure signals WAAAAY more sensitive than cardiac do not use Shunyata ... their advert of use in the medical field looks like purely a publicity stunt to me. 

[manueljenkin]

1 hour ago, jabbr said:

 

Did you hear? CERN just bought out the remaining stock of Shunyata, for use with the Large Hadron Collider ...

 

No seriously, EMI labs etc, and many which measure signals WAAAAY more sensitive than cardiac do not use Shunyata ... their advert of use in the medical field looks like purely a publicity stunt to me. 

I don't understand your issue. I asked them for knowledge transfer in an area I have a vague clue and they probably have a better picture of what's going on. What's wrong in that? I specified what I want to know and how it works to an extent they can share.

 

I don't think CERN will ask topping or smsl to design their components either, or ask ASR to validate their components with just static tones. Also CERN won't use consumerish components or even consumerish power supply/transmission lines. https://home.cern/science/engineering/powering-cern . I'd be happy to learn from them just the same way, just that they are not accessible to me currently. The shunyata and uptone guys reply to mails and, the latter have already replied to my mails. Whether I think something is valid or not I'll decide after consulting with professors and researchers (not online warriors).

 

I am already able to comprehend shunyata's pitch/demo I just wanted to know it in a deeper level. Maybe it relates to audio maybe it doesn't, I can't say without trying and I'm least bothered about audio when compared to my interest in other domains. Them having a medical division is just icing on top of the cake for me. Just an additional validation. Don't you see such things claimed by Apple, microsoft, Google etc. That their stuff is being used in mission critical jobs.

 

What's up with these so called objectivists dismissing anything of scientific concern or knowledge transfer.

[jabbr]

18 minutes ago, manueljenkin said:

I am already able to comprehend shunyata's pitch/demo I just wanted to know it in a deeper level. Maybe it relates to audio maybe it doesn't, I can't say without trying and I'm least bothered about audio when compared to my interest in other domains. Them having a medical division is just icing on top of the cake for me. Just an additional validation. Don't you see such things claimed by Apple, microsoft, Google etc. That their stuff is being used in mission critical jobs.

cool!

[pkane2001]

13 minutes ago, manueljenkin said:

or ask ASR to validate their components with just static tones. 

What's up with these so called objectivists dismissing anything of scientific concern or knowledge transfer.

 

Speaking of which. I know that you had mentioned that static tones are a problem before. Can you elaborate what the issue is, and how your test signal (I believe you mentioned you developed one) solves it. I'm genuinely curious, as I've been working on some test signals recently.

[manueljenkin]

6 minutes ago, pkane2001 said:

 

Speaking of which. I know that you had mentioned that static tones are a problem before. Can you elaborate what the issue is, and how your test signal (I believe you mentioned you developed one) solves it. I'm genuinely curious, as I've been working on some test signals recently.

I'd like to take more time, get it reviewed and published as a paper. I'll keep mum on details till then. Trying to ask guidance in most online forums was going nowhere in my experience and I'm not interested in going through that loop again. I've got enough leads and a few of my tests have been a success. Issue is simple, we assume everything to be LTI, which they are not, even the sampling process has aberrations and compensations for the same. So simple tests like autocorrelation from sine sweep can't parametrize them fully. A lot of things are still under study just like human perception.

 

I didn't say static tones are a "problem" when probed properly. I just said they are not a complete analysis of the device in hand. And you can't conclude anything without having a complete analysis/parametrization.

[jabbr]

4 hours ago, Superdad said:

And the perturbations we are looking at do not need to be very large to have the effect on clock threshold jitter that we believe are the root cause of the sonic differences heard.

I’m still waiting ... 

 

But since you’ve apparently given up “leakage current” and are now locked onto phase noise  these are all quite measurable things ... when you state your belief, that implies certain electrical behavior ... so when you say “they don’t have to be very large” you can apply noise to the clock oscillator power supply and measure phase noise. This has been done! The required noise is a very measurable!!!

 😂 

 

(and for the home listeners, I certainly believe that reducing phase noise at the DAC might improve SQ, but I don’t believe that phase noise worms it’s way from a server through a network and into a DAC unless there’s something broken  )

[jabbr]

12 minutes ago, manueljenkin said:

 

I didn't say static tones are a "problem" when probed properly. I just said they are not a complete analysis of the device in hand. And you can't conclude anything without having a complete analysis/parametrization.

Said another way: the behavior of a system given an arbitrary signal is not the linear sum to of the behaviors of the pure tones that make up the signal — said another way real systems display nonlinearity. 

[pkane2001]

19 minutes ago, manueljenkin said:

I'd like to take more time, get it reviewed and published as a paper. I'll keep mum on details till then. Issue is simple, we assume everything to be LTI, which they are not, even the sampling process has aberrations and compensations for the same. So simple tests like autocorrelation from sine sweep can't parametrize them fully. A lot of things are still under study just like human perception.

 

I didn't say static tones are a "problem" when probed properly. I just said they are not a complete analysis of the device in hand. And you can't conclude anything without having a complete analysis/parametrization.

 

Ok, since we can't talk about your test signal, maybe we can critique mine.  Multitone signal, auto-generated with up to many thousands of tones, low-crest optimized. Analysis below removes the test signal and leaves all distortion, shown in white. This includes HD, IMD, jitter, and all the noise.

 

The number on the right is an RMS value of total distortion plus noise. Example below is real, captured through a relatively inexpensive Apogee interface DAC/ADC loopback. Is this static? What doesn't this capture? What issues do you see?

 

(by the way, frequency response and phase are both easy to derive from the result, but I've yet to do it)

[jabbr]

48 minutes ago, manueljenkin said:

 

I don't think CERN will ask topping or smsl to design their components either, or ask ASR to validate their components with just static tones. Also CERN won't use consumerish components or even consumerish power supply/transmission lines. https://home.cern/science/engineering/powering-cern . I'd be happy to learn from them just the same way, just that they are not accessible to me currently.

Actually CERN makes its data publicly available so if you have the data analysis capabilities you can verify for yourself. Materials & methods are published so if you care to read ... a lot 

[manueljenkin]

4 minutes ago, pkane2001 said:

 

Ok, since we can't talk about your test signal, maybe we can critique mine.  Multitone signal, auto-generated with up to many thousands of tones, low-crest optimized. Analysis below removes the test signal and leaves all distortion, shown in white. This includes HD, IMD, jitter, and all noise.

 

The number on the right is an RMS value of total distortion plus noise. Example below is real, captured through an inexpensive Apogee interface DAC/ADC loopback. Is this static? What doesn't this capture? What issues do you see?

 

(by the way, frequency response and phase are both easy to derive from the result, but I've yet to do it)

I need a bit more detail on the sampling rate etc. To check on the sampling artefacts and filtering artefacts will need an analysis of the ADC in use.

 

But this is more of a static/averaged signal. Transient analysis will be one from an inertial frame of reference. The path from inertia to steady state denotes the transient curves. The type of filter/sampler design to measure transients and steady state phenomenon need not be same. Every design will have a bound within which they perform optimal.

[manueljenkin]

2 minutes ago, jabbr said:

Actually CERN makes its data publicly available so if you have the data analysis capabilities you can verify for yourself. Materials & methods are published so if you care to read ... a lot 

Thank you very much. You have given me some reading to do 💪.

[mikicasellas]

Have you seen this ?