A novel way to massively improve the SQ of computer audio streaming

[rickca]

59 minutes ago, lmitche said:

SATA SAGA update: Just tested the Amphenol 6" Sata cable against the 6" Coboc flat silver cable.  Coboc wins with a much fuller, dynamic sound.  99 cents on Egghead! 10" Coboc sounds good too!

So I shouldn't spend 229 euros on a Pachenko SATA Reference cable? 

[greenleo]

I doubt if anybody has done the direct  comparison yet.

[gadgetman]

6 hours ago, lmitche said:

Have you tried Daphile with a hard disk?  Just because an SSD isn't being accessed doesn't mean it isn't singing on the wire.

My original setup was Sandisk Ultra II 960G. Recently I changed it to Seagate ST2000LM007 because my music library is growing. I found the SSD sounds better than HDD. HDD sound a little bit muddy in my system. But the price difference didn't worth me to upgrade to a 2GB SSD. So I didn't buy one.

 

I'm using a fanless NUC. There's no fan in the system. HDD would be the only source of vibration. For Macmini users the fan is a necessity. The vibration is already there. Using HDD may or may not make it worse. So YMMV.

[lmitche]

10 hours ago, rickca said:

So I shouldn't spend 229 euros on a Pachenko SATA Reference cable? 

I have several SATA cables I can sell you for 458 euros each. They are twice the price, so twice as good, promise!

[HeeBroG]

On 27/03/2017 at 11:19 AM, romaz said:

 

I think the big X factor is the soon-to-be-released Iso Regen by Uptone.  I suspect this could be a better option than the Adnaco and should pair well with their LPS-1.

 

Is the upcoming tX-USBUltra from SOtM a similar product to the ISO Regen?

[HeeBroG]

On 15/04/2017 at 11:28 PM, gadgetman said:

Since Daphile has a RAM play function, if you have enough memory that you can load entire album into the memory. I would suggest to stick with SSD. My take is that after loading the album onto RAM there's no SSD accessing activity. So the electrical noise can be kept to minimum.

 

The second reason I don't recommend a HDD for Daphile is that there's always vibration from the HDD even though you put entire album into RAM. And vibration can degrade the SQ.

Thanks for the advice gadgetman

 

G

[greenleo]

3 hours ago, HeeBroG said:

Is the upcoming tX-USBUltra from SOtM a similar product to the ISO Regen?

Similar in terms of function, yes.  We're interested in their performances.

[mozes]

11 minutes ago, greenleo said:

Similar in terms of function, yes.  We're interested in their performances.

The tX doesn't have galvanic isolation

[greenleo]

2 hours ago, mozes said:

The tX doesn't have galvanic isolation

Correct but without galvanic isolation sounds better according to Sotm.  Hence the performance was more important.

 

 The major function of them was to reclock/regen the USB signal, hence similar function.

[sbilotta]

5 minutes ago, greenleo said:

Correct but without galvanic isolation sounds better according to Sotm.  Hence the performance was more important.

 

but is it not correct that if you power the tX with a LPS-1 you would nonetheless have galvanic isolation ?

[One and a half]

1 hour ago, greenleo said:

Correct but without galvanic isolation sounds better according to Sotm.  Hence the performance was more important.

 

 The major function of them was to reclock/regen the USB signal, hence similar function.

This means SOTM couldn't control the jitter when using galvanic isolation and used the Regen technology instead. Nothing wrong with that,  but the primary goal of stopping leakage currents from the computer to the DAC job is incomplete. 

[One and a half]

1 hour ago, sbilotta said:

but is it not correct that if you power the tX with a LPS-1 you would nonetheless have galvanic isolation ?

Mmm partially. Leakage currents from the power source are blocked,  but they still feed through from the source eg computer or media server.

[sbilotta]

15 minutes ago, One and a half said:

Mmm partially. Leakage currents from the power source are blocked,  but they still feed through from the source eg computer or media server.

ok, but if the source was fed by a LPS-1 too, would that block all leakage currents to the dac ?

[austinpop]

Several things:

1. SQ is improved both by improving signal integrity, and by squashing leakage loops.
2. Signal integrity can be improved by better clocking and regeneration. Regenerators like the RUR, Regen, iFi etc have been around for a while. In this thread @romaz, @mozes and others have described the effect of superior clocks using the sCLK-EX Ultra SOtM products.
3. Leakage loops can be broken by using isolated power supplies like the LPS-1 and VR Mini and/or by isolators in the signal path, like Intona for USB, Baaske/EMO EN-70HD/FMCs on Ethernet
4. As Alex has mentioned, and SOtM has communicated to me and Roy, implementing galvanic isolation in the signal path on USB is challenging in terms of jitter. Uptone has taken the approach of trying to mitigate this problem in their eagerly-awaited ISO-Regen. SOtM have taken a different tack - until they find a galvanic isolation solution that they feel has low enough jitter, they've chosen to refrain from addressing it.

Now here's the thing. There is no RIGHT ANSWER, nor is there a sure-fire cook book for newbies on offer. All of us are doing empirical, subjective, listening, and reporting our findings. YMMV! 

 

If you want my opinion - and it is just that, an opinion - it is preferable to address leakage loops primarily via isolated PSUs, which does not directly impact - and very likely, improves - signal integrity as well. Once you have that basic foundation, try isolation in the signal path - and see if it helps or hinders. 

 

For example - @mozes reported no improvement using Intona in his chain with tX-USBultra here:

However, others like @lmitche find benefit from EN-70HD isolators in their chain.

 

My own experience: I used to use FMCs upstream of my Aries Mini streamer. But since I went to:

• Direct connection using bridging
• sMS-200 powered by LPS-1

I've found the FMCs actually sound worse in the chain.

 

Like I said - wait for it - YMMV!

[mozes]

I am planning to try the Iso-Regen with my tX and then I can comment on the effect of GI based on empirical finding

[greenleo]

5 hours ago, sbilotta said:

but is it not correct that if you power the tX with a LPS-1 you would nonetheless have galvanic isolation ?

But this is not the function of the tx.  You may power it by a battery and get the galvanic isolation as well.

[greenleo]

4 hours ago, One and a half said:

This means SOTM couldn't control the jitter when using galvanic isolation and used the Regen technology instead. Nothing wrong with that,  but the primary goal of stopping leakage currents from the computer to the DAC job is incomplete. 

I would call it the best balance SOTM found or taken.  I am not defending SOTM but taking any assembled components, like a CAS, has to achieve lots of things and has intended systems for application as well.  Hence balancing becomes very important.  Suppose your DAC is galvinic isolated, then let go of the galvinic isolation becomes a better strategy and if better SQ is provided, then the product wins.  "The primary goal of stopping leakage current ….." may not be that important in some system, then the product wins.

 

We don't have a product that is "perfect" in terms of it sounds better in EVERY system, size small enough so that it may be applied in every system AND is the cheapest.  If so, other product products will become obsolete.

 

Finally, I agree with the statement "SOTM couldn't control ..." and I am saying this is not necessarily a bad thing.  At the end only SQ ccounts.

 

My 2 cents.

[mozes]

1 hour ago, greenleo said:

I would call it the best balance SOTM found or taken.  I am not defending SOTM but taking any assembled components, like a CAS, has to achieve lots of things and has intended systems for application as well.  Hence balancing becomes very important.  Suppose your DAC is galvinic isolated, then let go of the galvinic isolation becomes a better strategy and if better SQ is provided, then the product wins.  "The primary goal of stopping leakage current ….." may not be that important in some system, then the product wins.

 

We don't have a product that is "perfect" in terms of it sounds better in EVERY system, size small enough so that it may be applied in every system AND is the cheapest.  If so, other product products will become obsolete.

 

Finally, I agree with the statement "SOTM couldn't control ..." and I am saying this is not necessarily a bad thing.  At the end only SQ ccounts.

 

My 2 cents.

Agree with your argument, SQ is the ultimate goal, doesn't matter much how to get there with or without GI. 

[romaz]

6 hours ago, sbilotta said:

ok, but if the source was fed by a LPS-1 too, would that block all leakage currents to the dac ?

 

Proper galvanic isolation does much more than just block leakage current introduced by power supplies or address ground loops.  In my system, RF noise in the ground plane that is generated by the source component itself is much more prominent.

 

I have done extensive testing with battery supplies (not the LPS-1) in the past and compared against some of my lesser supplies like my iFI, HDPlex, Teradak, etc., even though my battery supplies have no leakage current, they never sounded that much better and in some cases, they sounded worse.  At least in these instances, leakage current wasn't a big deal.

 

When my LPS-1 arrived, I compared it against my other battery supplies and while leakage current wasn't an issue with any of these supplies, my LPS-1 soundly trounced all the battery supplies I had on hand.  In my system, it was clear to me that the strength of the LPS-1 was due more to its very low output impedance than its lack of leakage current.

 

I don't want to downplay or diminish the significance of leakage current because in some systems, it could be much more significant than in mine but once again, here are John's own words about the significance of leakage current relative to "other" things:

 

"Leakage current has always been there, but has not been a big issue until recently for two reasons:

 

A) it is fairly low level in most systems so its effects have been masked by all the other problems preventing our systems from being their best, as these other sources of problems have been identified and addressed, the sonic results of the leakage current are now much more obvious."

 

Minimizing RF noise (which is not the same thing as leakage current) should be the real target of galvanic isolation.  With some of my server builds in the past, I had gone to great lengths to minimize RF with such practices as using thick aluminum enclosures, using well insulated DC cabling and making sure no bare wire was exposed, using point-to-point wiring to keep wire lengths to a minimum, underclocking my CPU and RAM, independently powering my OS and storage drives, shielding my CPU, wrapping my cabling, RAM and SSDs in ERS paper, and even lining the ICs on my motherboard as well as the entire inside of my chassis with ERS paper and despite these painstaking measures, I found it astonishing that when I grounded the signal from my server with an Entreq grounding box, noise floor dramatically dropped further.  With the Entreq in place, I would then tell myself "finally, RF is now gone and I now have the lowest noise floor achievable..."  But then, the next new thing comes out and I realize that my server's noise floor can drop even lower.  RF is an incredibly pesky thing and, in my experience, difficult, if not impossible to completely get rid of.  

 

Galvanic isolation, however, doesn't come without its problems.  As we know, Toslink has the ability to completely block ground plane RF but it is plagued with jitter issues.  SOtM reportedly also did testing with GI devices with the desire, I'm sure, of wanting to incorporate the benefits of GI within their own devices, however, what they supposedly found with such devices including the Intona was unacceptable amounts of jitter and so, at this time, they have intentionally chosen not to incorporate galvanic isolation into any of their devices.  As a clock-oriented company, I guess I'm not surprised.

 

Not to say that attempting to achieve GI is a mistake, it just has to be implemented well.  Based on my own comparisons, my Chord DAVE, I believe, is a good example of this.  Rob Watts went to great lengths to "float" his USB input so that it would be on par with the RF immunity of the DAVE's optical input.  With Hugo, Mojo and all of Chord's battery-operated DACs, because their USB inputs aren't galvanically isolated, their optical inputs sound best and Rob believes this as well.  With DAVE, comparing the optical vs USB side by side, USB sounds a touch brighter but with this DAC, USB reigns supreme compared all other inputs for a variety of reasons although Rob has suggested DAVE's USB's superiority is in large part due to the effectiveness of its galvanic isolation.  

 

I suspect the Uptone Iso Regen wouldn't be coming to market if John and Alex didn't believe they had figured out a way to properly implement galvanic isolation and so I know many of us eagerly await its release.  As good as I suspect it will be (and I'm confident it will be great given their track record), I still don't believe it will be a complete fix against RF.  In the same way that my observations have shown me that one very good clock can't repair the damage from a multitude of bad clocks before it, I believe the same thing applies to GI.  Even if the Iso Regen effectively eliminates all RF noise in the ground plane, it cannot do anything about the RF noise that has permanently embedded itself into the actual signal, at least that's my theory.  If this was true, than none of the upstream measures I have been taking these past few months should have made a lick of difference to my supposedly RF/jitter immune DAC and yet, the improvements have been remarkable.  I believe to properly address RF (and jitter), you have to address it at every step.

[greenleo]

55 minutes ago, romaz said:

 

Proper galvanic isolation does much more than just block leakage current introduced by power supplies or address ground loops.  In my system, RF noise in the ground plane that is generated by the source component itself is much more prominent.

 

I have done extensive testing with battery supplies (not the LPS-1) in the past and compared against some of my lesser supplies like my iFI, HDPlex, Teradak, etc., even though my battery supplies have no leakage current, they never sounded that much better and in some cases, they sounded worse.  At least in these instances, leakage current wasn't a big deal.

 

When my LPS-1 arrived, I compared it against my other battery supplies and while leakage current wasn't an issue with any of these supplies, my LPS-1 soundly trounced all the battery supplies I had on hand.  In my system, it was clear to me that the strength of the LPS-1 was due more to its very low output impedance than its lack of leakage current.

 

I don't want to downplay or diminish the significance of leakage current because in some systems, it could be much more significant than in mine but once again, here are John's own words about the significance of leakage current relative to "other" things:

 

"Leakage current has always been there, but has not been a big issue until recently for two reasons:

 

A) it is fairly low level in most systems so its effects have been masked by all the other problems preventing our systems from being their best, as these other sources of problems have been identified and addressed, the sonic results of the leakage current are now much more obvious."

 

Minimizing RF noise (which is not the same thing as leakage current) should be the real target of galvanic isolation.  With some of my server builds in the past, I had gone to great lengths to minimize RF with such practices as using thick aluminum enclosures, using well insulated DC cabling and making sure no bare wire was exposed, using point-to-point wiring to keep wire lengths to a minimum, underclocking my CPU and RAM, independently powering my OS and storage drives, shielding my CPU, wrapping my cabling, RAM and SSDs in ERS paper, and even lining the ICs on my motherboard as well as the entire inside of my chassis with ERS paper and despite these painstaking measures, I found it astonishing that when I grounded the signal from my server with an Entreq grounding box, noise floor dramatically dropped further.  With the Entreq in place, I would then tell myself "finally, RF is now gone and I now have the lowest noise floor achievable..."  But then, the next new thing comes out and I realize that my server's noise floor can drop even lower.  RF is an incredibly pesky thing and, in my experience, difficult, if not impossible to completely get rid of.  

 

Galvanic isolation, however, doesn't come without its problems.  As we know, Toslink has the ability to completely block ground plane RF but it is plagued with jitter issues.  SOtM reportedly also did testing with GI devices with the desire, I'm sure, of wanting to incorporate the benefits of GI within their own devices, however, what they supposedly found with such devices including the Intona was unacceptable amounts of jitter and so, at this time, they have intentionally chosen not to incorporate galvanic isolation into any of their devices.  As a clock-oriented company, I guess I'm not surprised.

 

Not to say that attempting to achieve GI is a mistake, it just has to be implemented well.  Based on my own comparisons, my Chord DAVE, I believe, is a good example of this.  Rob Watts went to great lengths to "float" his USB input so that it would be on par with the RF immunity of the DAVE's optical input.  With Hugo, Mojo and all of Chord's battery-operated DACs, because their USB inputs aren't galvanically isolated, their optical inputs sound best and Rob believes this as well.  With DAVE, comparing the optical vs USB side by side, USB sounds a touch brighter but with this DAC, USB reigns supreme compared all other inputs for a variety of reasons although Rob has suggested DAVE's USB's superiority is in large part due to the effectiveness of its galvanic isolation.  

 

I suspect the Uptone Iso Regen wouldn't be coming to market if John and Alex didn't believe they had figured out a way to properly implement galvanic isolation and so I know many of us eagerly await its release.  As good as I suspect it will be (and I'm confident it will be great given their track record), I still don't believe it will be a complete fix against RF.  In the same way that my observations have shown me that one very good clock can't repair the damage from a multitude of bad clocks before it, I believe the same thing applies to GI.  Even if the Iso Regen effectively eliminates all RF noise in the ground plane, it cannot do anything about the RF noise that has permanently embedded itself into the actual signal, at least that's my theory.  If this was true, than none of the upstream measures I have been taking these past few months should have made a lick of difference to my supposedly RF/jitter immune DAC and yet, the improvements have been remarkable.  I believe to properly address RF (and jitter), you have to address it at every step.

Hi Roy,

 

Very informative and interesting.  Thank you.

 

Do you imply that even use Battery to power the server, the Entreq grounding box still help?  If so, which box are you using?

 

The elimination of RF from the grounding box is higher than the sum of the rest of your painstaking measures?

 

How effective did you found in underclocking the RAM and the CPU?

 

You mentioned point to point connection.  However, it seems that the Pachanko SATA cable has minimum length to achieve better SQ.  Any explanations/thoughts?

[romaz]

10 minutes ago, greenleo said:

Do you imply that even use Battery to power the server, the Entreq grounding box still help?  If so, which box are you using?

 

No, I have never owned a battery supply that has been capable of powering a server.  Vinnie Rossi, however, just came out with a 15-24V version of his MINI PURE-DC-4EVR supercap PSU that can output up to 5A (peak) that could possibly power some lightweight servers.  

 

As I stated in my post, we are talking about small devices like the microRendu, FMCs, and switches and these devices are not always easy to signal ground.  While I have not tried to battery power my router, connecting my router to my grounding box via its USB input resulted in a small but easily noticeable improvement.  Electronic equipment that are very sensitive to the impact of RF (like preamps) or create considerable RF themselves (like computers) respond very nicely to grounding boxes.  I used to own the Entreq Posdeidon along with a mix of Atlantis grounding cables.  For convenience reasons, I have moved on to the Synergistic Research Grounding Box.

 

13 minutes ago, greenleo said:

The elimination of RF from the grounding box is higher than the sum of the rest of your painstaking measures?

 

No, my point was that after all of my previous painstaking measures, which resulted in quite a bit of improvement collectively, adding the grounding box resulted in further significant improvement which I didn't think was possible given everything I had already done.  I think we all know what it's like to buy or build something and believe it's the best there is, or at least we believe there's no way it can be better than this...at least until the next "better" thing comes along.  Same with RF, just when we think we've hit rock bottom with respect to noise floor due to all of our RF-reducing efforts, something else comes along that allows us to reduce RF (and, therefore, noise floor) even further.  As long as we have to contend with these noisy off-the-shelf motherboards, I don't believe we will ever be able to hit rock bottom with RF.

 

24 minutes ago, greenleo said:

How effective did you found in underclocking the RAM and the CPU?

 

The difference was small but still very desirable and because it costs nothing to implement, it was very worthwhile.  On the whole, the difference was more noticeable than the even smaller improvements I heard from some of my battery supplies.

 

26 minutes ago, greenleo said:

You mentioned point to point connection.  However, it seems that the Pachanko SATA cable has minimum length to achieve better SQ.  Any explanations/thoughts?

 

I think you misunderstand what "point-to-point" wiring is.  These days, there's less point-to-point wiring going on due to the dominance of PCBs but open up any computer and you'll see a mess of wires, nonetheless.  Many computer builders will route their wires through conduit or along the walls of the chassis to allow for a "cleaner" look but to do this often leads to longer runs of wire.  Moreover, to run wires with other wires just to achieve this "cleaner" look will also increase the likelihood of crosstalk resulting in increased noise.  With point-to-point wiring, generally, the goal is to connect two electronic components together as directly as possible (with as short a run of wire as possible).  The end result isn't always pretty but many will tell you that SQ is better this way.  Here is an ugly example of point-to-point wiring:

 

 

As for my Pachanko SATA cable, this cable was cut to my specified length (20cm) and will go directly from drive to SATA port.  This is a form of point-to-point wiring.

 

 

[baconbrain]

Perhaps it is just me, but hasn't this thread really grown into a bit of monster ...?

 

Not sure if possible, but with all of the fantastic info/discussions contained within, I personally think it would be very helpful if it could be broken apart into dedicated sub-topics, etc...  

[romaz]

17 hours ago, baconbrain said:

Perhaps it is just me, but hasn't this thread really grown into a bit of monster ...?

 

Not sure if possible, but with all of the fantastic info/discussions contained within, I personally think it would be very helpful if it could be broken apart into dedicated sub-topics, etc...  

 

 

I agree.  As I find myself looking for a certain post by a certain individual, leafing through nearly 1,500 posts over 58 pages is no small task.  While I meant it to be a discussion of a novel way to improve either the microRendu or sMS-200, it has unintentionally evolved from this direct connection between server and endpoint via bridged LAN ports to power supplies, clocking, ideal music server configurations, optimized OS solutions, motherboards, ideal storage media, SATA cables, Ethernet cables, endpoints other than the mR or sMS-200, and galvanic isolation.  

 

For those just jumping in, I can imagine that this thread looks like one big disjointed mess but as someone who has been a part of it from the beginning, and speaking only for myself, it has been a wonderful and stimulating think tank where innovative ideas and experiences have been shared, discussed, affirmed and challenged.  Nothing advances progress better than a broad community effort and without a doubt, my server build would look vastly different without the input and inspiration provided by so many.    

 

If all goes well, my reclocked motherboard will arrive from SOtM sometime next week.  At that point, I will report on the impact of removing as many bad clocks as possible from the server and whether this negates or further enhances the impact of endpoints like the mR or sMS-200.  Despite the broad topics that have been discussed, for me, it has still always been about finding ways to improve upon the mR or sMS-200.  Shortly after that, I will be making my exit from this thread (and from posting on forums, in general), at least for the foreseeable future due to other more pressing and time-consuming commitments.  Because we are talking about a few weeks of time, I will likely make no attempt to start any new threads although I certainly won't discourage anyone from breaking this apart into dedicated sub-topics.  It's been a fun ride...

[rickca]

5 minutes ago, romaz said:

 

 

If all goes well, my reclocked motherboard will arrive from SOtM sometime next week.  At that point, I will report on the impact of removing as many bad clocks as possible from the server and whether this negates or further enhances the impact of endpoints like the mR or sMS-200.  

Your contribution has been invaluable.  Sorry to hear you need to disengage but of course I understand.  I'm really looking forward to your assessment of the clock upgrades.

[baconbrain]

3 hours ago, romaz said:

If all goes well, my reclocked motherboard will arrive from SOtM sometime next week.  At that point, I will report on the impact of removing as many bad clocks as possible from the server and whether this negates or further enhances the impact of endpoints like the mR or sMS-200.  Despite the broad topics that have been discussed, for me, it has still always been about finding ways to improve upon the mR or sMS-200.  Shortly after that, I will be making my exit from this thread (and from posting on forums, in general), at least for the foreseeable future due to other more pressing and time-consuming commitments.  Because we are talking about a few weeks of time, I will likely make no attempt to start any new threads although I certainly won't discourage anyone from breaking this apart into dedicated sub-topics.  It's been a fun ride...

Hope I didn't toss the rock that started the landslide ... 

 

For me, this has been one of the most informative threads that I have read in a long time and really appreciate your contributions. Looking forward to your summary, and return to CA, when time permits.