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

[Nenon]

13 hours ago, lateboomer said:

Another thing is for single PC setup running Euphony Stylus, if os should be run in Optane card then where should music files be stored in optimized way?  Is it ssd or hdd? What is the type. I tried to skim through the index but cannot find the info. 

 

I started a topic on this. Haven't updated the index in a while but will check if there are any new findings I can add.

 

On 3/26/2019 at 7:38 PM, Nenon said:

This would be an attempt to extract some information from the “A novel way to massively improve the SQ of computer audio streaming” thread related to computer storage. 

 

People are trying different methods. Here are some of the options:

1. SSD drive 

2. HDD drive 

3. External HDD with external power supply

4. Optane memory

5. NAS accessed over the network

6. Run everything in ramroot

 

Let’s discuss each of these. 

 

1. SSD drive

The general consensus is that SSD is noisy and should not be used. Powering by LPS helps but does not eliminate the noise. 

 

On the other hand, many music servers use SSD including Innuos Statement.

 

 

2. HDD Drive

Some people say it's better than SSD, others say it's worse. I will try to extract some posts on this topic later.

 

 

 

3. External HDD with external power supply

Keeps the HDD vibration out of your chassis and does not pollute your ATX power.

 

Seems like the quality of cables and power supply makes a difference.

 

 

 

 

 

4. Optane memory

Very good feedback so far. The problem it has less memory than HDD, or SSD, or SAN, which makes it not very suitable for a huge collection of ripped CDs. Very good option for OS booting. It's also not supported on every motherboard. 

 

 

5. NAS accessed over the network

Good results reported and prefered method to some of the leading members in this forum. It probably makes sense to combine with good network cards and switch(es). 

 

 

 

6. Run everything in ramroot

Not even sure if this option should be here. It probably belongs to the AL thread. 

 

 

 

Other interesting posts below.

 

 

 

 

[Triplefun]

I just joined this site and am overwhelmed by the wealth of information I have also been following https://www.stereo.net.au/forums/topic/82802-building-the-idealish-music-server/page/316/ which you may also find to be of interest.

[rickca]

Looks like there is a downside to the new x570 motherboards supporting PCIe V4.  The chipset has a fan!  Apparently, the chipset is 15W compared to 5W on x470.  

https://www.pcgamesn.com/amd/amd-x570-chipset-fan-nobody-wants-this

[cool_chris]

I am impressed by the knolage of Romaz and Austinpop.

I am scientist and part time audiopile. Was building my servers for many years and 

latency was always my „drug” to the point I was unable to boot the optimised W12server anymore.

I was modifying my W12server root , motherboard and thausends of other things.

I quit few years ago due to the unstability of the system.

 

I know the field has changed a lot. It is impressive how much more we know this days.

Could you please think and post the ultimate server / endpoint solution ?

 

I was thinking to get the Taiko Extreme but wondering if this will be the best for the totaldac tvelve system ?

 

for that kind of money I could think of building the optimum system myself based on Euphony ?

what parts will be the best setup in your opinion ?

no ssd ! Optane for music library and system ?

I  say cost no object for this build.

 

[The Computer Audiophile]

3 minutes ago, cool_chris said:

I am impressed by the knolage of Romaz and Austinpop.

I am scientist and part time audiopile. Was building my servers for many years and 

latency was always my „drug” to the point I was unable to boot the optimised W12server anymore.

I was modifying my W12server root , motherboard and thausends of other things.

I quit few years ago due to the unstability of the system.

 

I know the field has changed a lot. It is impressive how much more we know this days.

Could you please think and post the ultimate server / endpoint solution ?

 

I was thinking to get the Taiko Extreme but wondering if this will be the best for the totaldac tvelve system ?

 

for that kind of money I could think of building the optimum system myself based on Euphony ?

what parts will be the best setup in your opinion ?

no ssd ! Optane for music library and system ?

I  say cost no object for this build.

 

Any guy who goes by Cool Chris is OK in my book.

[kravi4ka]

Haha, Cool Chris is great indeed.

 

Well, I am trying to follow all threads, to read every single post and not miss anything. There are several people here who are to be loved for sharing it all, have not seen anything like that for many years.

 

But what I decided to do in the end is just do what bobfa did and just copy his NUC design. Even if it is not the BEST there is it sure is something I CAN do. FOMO is a serious medical condition now, might easily kill the whole purpose of listening to music, be careful.

[rickca]

For you Optane fans, Intel has officially announced the M15.  The M.2 card is now PCIe 3.0 x4 rather than x2.  Expected availability is middle of Q3.

https://www.anandtech.com/show/14437/intel-announces-optane-memory-m15-3d-xpoint-on-m2-pcie-30-x4 

 

Intel also announced Ice Lake CPUs built on 10nm process with 9W (Y-series), 15W and 28W TDP variants up to quad core.  They have built-in support for Thunderbolt 3 and Wifi 6.  

https://www.anandtech.com/show/14436/intel-10th-gen-10nm-ice-lake-cpus

[Nenon]

Come on guys... You can't read 567 pages with all the external references?!?

 

I would start by reading the banner above. It sums it up pretty good. What was considered "best" last month is different than what we are looking at today. And what we are looking at today would most likely be outdated by the end of the summer. 

 

What bobfa is trying to do is pretty good and easy to follow. We thank him for his efforts. However, it is not the bleeding edge technology people here are experimenting with. 

 

A cost no object computer, based on everything I’ve read here so far, in my opinion would be:
Motherboard: ASRock Motherboard (Z390 Phantom Gaming-ITX/AC)
CPU: i9-9900K
Powered by multiple rails Paul Hynes SR7 LPS. His silver DC cables are probably some of the best as well. Keep them as short as possible.
Euphony running on an Optane card.
High performance fanless chassis, such as the HDPlex H3 or the top Streacom. 
PinkFaun Ultra OXCO USB card - powered by 5V DC LPS (SR7 preferably)
JCAT NET Femto Card - powered by 5V DC LPS (SR7 preferably)

Get a couple of SOTM switches, good power supplies, and cables, and you can host your music on a NAS if it does not fit on the Optane drive. Vibration treatment, power cables, USB cables, fancy fuses, and all the jazz applies here.

 

Now, let's cover some problems and compromises:
1. The motherboard listed above has only one PCI-e slot, so you can’t fit both cards. One of the cards would have to be connected to the M.2 via questionable cables/convertors or you can use the Thunderbolt solution described here. Or you can try a motherboard that has more PCI-e slots, but there is no guarantee it would sound good, and you would run into problems finding suitable chassis that can host two PCI-e cards without questionable quality cable extenders.

2. Paul Hynes does not accept orders for SR7 power supplies anymore. 

3. Given that a SR7 is hard to obtain, one may consider a lower TDP CPU,  such as the i9900 or maybe even a i9900T. But how that would impact the sound quality is yet to be seen.
 

In my subjective opinion, I believe that's the latest trend as of the end of May. June is coming in a couple of days :). 

[bobfa]

12 minutes ago, Nenon said:

Come on guys... You can't read 567 pages with all the external references?!?

 

I would start by reading the banner above. It sums it up pretty good. What was considered "best" last month is different than what we are looking at today. And what we are looking at today would most likely be outdated by the end of the summer. 

 

What bobfa is trying to do is pretty good and easy to follow. We thank him for his efforts. However, it is not the bleeding edge technology people here are experimenting with. 

 

A cost no object computer, based on everything I’ve read here so far, in my opinion would be:
Motherboard: ASRock Motherboard (Z390 Phantom Gaming-ITX/AC)
CPU: i9-9900K
Powered by multiple rails Paul Hynes SR7 LPS. His silver DC cables are probably some of the best as well. Keep them as short as possible.
Euphony running on an Optane card.
High performance fanless chassis, such as the HDPlex H3 or the top Streacom. 
PinkFaun Ultra OXCO USB card - powered by 5V DC LPS (SR7 preferably)
JCAT NET Femto Card - powered by 5V DC LPS (SR7 preferably)

Get a couple of SOTM switches, good power supplies, and cables, and you can host your music on a NAS if it does not fit on the Optane drive. Vibration treatment, power cables, USB cables, fancy fuses, and all the jazz applies here.

 

Now, let's cover some problems and compromises:
1. The motherboard listed above has only one PCI-e slot, so you can’t fit both cards. One of the cards would have to be connected to the M.2 via questionable cables/convertors or you can use the Thunderbolt solution described here. Or you can try a motherboard that has more PCI-e slots, but there is no guarantee it would sound good, and you would run into problems finding suitable chassis that can host two PCI-e cards without questionable quality cable extenders.

2. Paul Hynes does not accept orders for SR7 power supplies anymore. 

3. Given that a SR7 is hard to obtain, one may consider a lower TDP CPU,  such as the i9900 or maybe even a i9900T. But how that would impact the sound quality is yet to be seen.
 

In my subjective opinion, I believe that's the latest trend as of the end of May. June is coming in a couple of days :). 

Thank you for the kind words!  I hate the bleeding edge, I pass out when I see blood.

 

I would build two of the above and then purchase a third board to send to SoTM for new clocks.  Then you would have duplicate endpoint and server to run together

 

NOTE I still think Xeon is a key component we are not taking into account yet. 

 

[austinpop]

7 hours ago, The Computer Audiophile said:

Any guy who goes by Cool Chris is OK in my book.

 

Even more impressive is he joined in 2012, and this is his first post. That is the epitome of cool-ness!

 

7 hours ago, cool_chris said:

I am impressed by the knolage of Romaz and Austinpop.

I am scientist and part time audiopile. Was building my servers for many years and 

latency was always my „drug” to the point I was unable to boot the optimised W12server anymore.

I was modifying my W12server root , motherboard and thausends of other things.

I quit few years ago due to the unstability of the system.

 

I know the field has changed a lot. It is impressive how much more we know this days.

Could you please think and post the ultimate server / endpoint solution ?

 

I was thinking to get the Taiko Extreme but wondering if this will be the best for the totaldac tvelve system ?

 

for that kind of money I could think of building the optimum system myself based on Euphony ?

what parts will be the best setup in your opinion ?

no ssd ! Optane for music library and system ?

I  say cost no object for this build.

 

 

Hi Cool Chris,

 

There is no single answer. If there were, we'd have no reason to keep experimenting. Strictly from my own experiences (and I suspect @romaz would agree with most of these), here are my thoughts.

1. Power supply quality is paramount, and this is usually the gating factor for most people. Sadly, we live in a world where the best available outboard PSU (Paul Hynes SR7) is a vanishingly scarce commodity. I wish we had other comparable choices. There are some emerging alternatives that show promise, but it remains to be seen if any can achieve the lofty heights of the SR7.
2. It's not just one PSU. When you add it all up, between ATX (3-rails), EPS (1 rail), Ethernet and USB cards (2 rails), you're already in need of 6 rails, and we haven't even talked about other devices like switches, USB reclockers etc. And some of these rails (ATX and EPS) require hefty current as well. This is what trips most people up. It's very difficult to get 6-8 rails of superlative PSU quality. This is why for most people, the ideal solution depends on the available power supplies.
3. Clock quality is important, and can have a transformative impact on refinement and resolution. However, there appears to be a very strong cross-correlation between PSU quality and clock quality. Based on my experience so far, if you can score excellent PSUs, the additional impact of clocking is actually lower. In other words, clock improvements and PSU improvements are not strictly additive. Further I would say, PSU quality matters far more than clock quality.
4. Faster CPUs with more cores and bigger caches have a huge impact on SQ - mostly in terms of dynamics and realism - even when running no DSP or upsampling. This is why the NUC7i7 sounds better than the CJYH/PJYH, and why the standalone server with i7-8700K sounds better still. This trumps the clock improvements that come from modding the NUC with sCLK-EX external clocking. Of course you have to pay for this in PSU rails. If your PSU quality is not sufficiently good, then you can only achieve better dynamics at the expense of harshness. This is why in many cases, if you only have a few low-current but otherwise outstanding PSUs like the LPS-1.2 or SR4 available, your best choice may still be to use a standalone NUC, or a pair of NUCs (server-endpoint). Yes, you give up some dynamics, but you get smoothness and refinement.
5. Finally, software still seems to matter. Between differences in OS (Windows, AudioLinux, Euphony etc), players (Stylus, Roon, HQPlayer), it is possible to achieve significant improvement after you've dialed in the HW.

This is why I say there is no one right answer. 

 

Also, I don't agree with the implication from an earlier poster that we flit from one topology to another in a whimsical manner. I see this as an expanding wave front of discovery. In a previous post I called it a furthering of the high water mark. Every time we achieve a new high, we find ways to improve it further. I've said this before and will say it again - just because a new finding expands the wavefront, or sets a new high-water mark, that does not negate the previous high! Unless you want to constantly be on the bleeding edge, you can rest a while (a long while!) at previous high points. This is why there are people who are still perfectly happy running with sMS-200ultra's, ZENith servers, NUCs and modded NUCs. 

 

We can all explore at our own pace. And there is value in letting others attack the bleeding edge, and learn from their efforts!

[cool_chris]

1 hour ago, austinpop said:

Even more impressive is he joined in 2012, and this is his first post. That is the epitome of cool-ness! 

 

It took me 7 years to read it all 😄

I am slow and You are posting so many things a day It was almost impossible to cach up with everything. 

 

Again. I did many things back in 2010 -2015 and tested all the options.

I had to quit because of to much time I burned to keep reinstalling everything after the system was not booting anymore. I had addition to the short latency.

it looks like I was right to go this direction , but hardwere could not keep the owners needs.

After I quit I purchesd totaldac server + twelve DAC and stayed with that till now.

Unfortunate that perfect sound I was getting old times is still on the back of my head . 

I have totaldac top DAC so no need for upsample.

 

After reading all of the topicks here I think it is a good time to get back and 

1. Purchese top server soch as Taiko Extreme 

( I wish this is linux based server not windows that I hate, because w12 server plays well once installed and optimised but it starts to DEGRADE the sound with time.

Reinstalling everything brings the SQ back but it takes a day or two to install and optimise everything.

please correct me if I am wrong.

 

OR

 

2. Build the cost no object server (euphony preferable) that WILL NOT turn down or degrade the SQ with time.

 

Aditionally 

I should add the Sotm swiches ( 2 pieces ) or wait for the Uptone ether regen to come.

[lmitche]

40 minutes ago, cool_chris said:

 

It took me 7 years to read it all 😄

I am slow and You are posting so many things a day It was almost impossible to cach up with everything. 

 

Again. I did many things back in 2010 -2015 and tested all the options.

I had to quit because of to much time I burned to keep reinstalling everything after the system was not booting anymore. I had addition to the short latency.

it looks like I was right to go this direction , but hardwere could not keep the owners needs.

After I quit I purchesd totaldac server + twelve DAC and stayed with that till now.

Unfortunate that perfect sound I was getting old times is still on the back of my head . 

I have totaldac top DAC so no need for upsample.

 

After reading all of the topicks here I think it is a good time to get back and 

1. Purchese top server soch as Taiko Extreme 

( I wish this is linux based server not windows that I hate, because w12 server plays well once installed and optimised but it starts to DEGRADE the sound with time.

Reinstalling everything brings the SQ back but it takes a day or two to install and optimise everything.

please correct me if I am wrong.

 

OR

 

2. Build the cost no object server (euphony preferable) that WILL NOT turn down or degrade the SQ with time.

 

Aditionally 

I should add the Sotm swiches ( 2 pieces ) or wait for the Uptone ether regen to come.

Get the Taiko Audio SGM Extreme. It is the best of the best. Did you know the Totaldac D1-twelve mk2 is one of the DACs used when testing the SQ of the Extreme?

 

While you are at it, checkout the Taiko Audio Daiza anti-vibration platforms. I have three here and the Panzerholtz construction has an immediate impact enriching SQ. My DAC, Amp and NUC/ISO Regen ride on these platforms. 

 

TheTaiko Audio guys are the top players in computer audio.

[gdpr]

Alternatively, you can buy the 'best' linux server/streamer Pinkfaun 2.16

Dirk

[cool_chris]

3 hours ago, austinpop said:

• Power supply quality is paramount, and this is usually the gating factor for most people. Sadly, we live in a world where the best available outboard PSU (Paul Hynes SR7) is a vanishingly scarce commodity. I wish we had other comparable choices. There are some emerging alternatives that show promise, but it remains to be seen if any can achieve the lofty heights of the SR7.

 

 

you might be surprised but I cocected a lot of supplys over this years.

I own 6  PH SR7, 3 Bakoons, 3 Uptone 1.2 and 4 Vinnie Rossi mini 5 and 12v.

I also have 800w custom  linearvATX supply by core audio from old days that I keep in bacement for last 5 years.

I can use all this so not much limitations here.

3 hours ago, austinpop said:

• It's not just one PSU. When you add it all up, between ATX (3-rails), EPS (1 rail), Ethernet and USB cards (2 rails), you're already in need of 6 rails, and we haven't even talked about other devices like switches, USB reclockers etc. And some of these rails (ATX and EPS) require hefty current as well. This is what trips most people up. It's very difficult to get 6-8 rails of superlative PSU quality. This is why for most people, the ideal solution depends on the available power supplies.

 

as stated above I have a big collection of all kinds of supplys so not much limitations here.

3 hours ago, austinpop said:

 

• Clock quality is important, and can have a transformative impact on refinement and resolution. However, there appears to be a very strong cross-correlation between PSU quality and clock quality. Based on my experience so far, if you can score excellent PSUs, the additional impact of clocking is actually lower. In other words, clock improvements and PSU improvements are not strictly additive. Further I would say, PSU quality matters far more than clock quality.

I got this and fully agree with you. This is why the Taiko went to the double processor with 10 cores each with No upsampling. Powering everything well.

they also avoided the oxco clocks modes and they use stock board clock !

Powering the clock with clean current is far more important that its quality .

low phase noise is what gives SQ

3 hours ago, austinpop said:

 

• Faster CPUs with more cores and bigger caches have a huge impact on SQ - mostly in terms of dynamics and realism - even when running no DSP or upsampling. This is why the NUC7i7 sounds better than the CJYH/PJYH, and why the standalone server with i7-8700K sounds better still. This trumps the clock improvements that come from modding the NUC with sCLK-EX external clocking. Of course you have to pay for this in PSU rails. If your PSU quality is not sufficiently good, then you can only achieve better dynamics at the expense of harshness. This is why in many cases, if you only have a few low-current but otherwise outstanding PSUs like the LPS-1.2 or SR4 available, your best choice may still be to use a standalone NUC, or a pair of NUCs (server-endpoint). Yes, you give up some dynamics, but you get smoothness and refinement.

This is why the Taiko went to the double processor with 10 cores each with No upsampling I think.

They also avoided the oxco clocks modes and they use stock clock.

Their construction looks like it is double box in one box.

they dedicated each CPU with part of the work ( 1 renderer and 1 streemer).

is this taiko extreme better than fully moded custom server ? I dont know.

3 hours ago, austinpop said:

• Finally, software still seems to matter. Between differences in OS (Windows, AudioLinux, Euphony etc), players (Stylus, Roon, HQPlayer), it is possible to achieve significant improvement after you've dialed in the HW.

 

yes. I just have good and bad experience at the same time with W12 server.

it sounded excellent but short time. I hate W OS because it ruins my setup with time and degrades the SQ. I hope Euphony might be a good solution.

[cool_chris]

53 minutes ago, ddetaey said:

Alternatively, you can buy the 'best' linux server/streamer Pinkfaun 2.16

Dirk

You mean the Pinkfaun 2,16x with all the optional things.

  And 2 pieces . One renderer and one for streeming.

also if you add all the options that work you will end with 33 - 34 K Euro.

[bibo01]

32 minutes ago, cool_chris said:

This is why the Taiko went to the double processor with 10 cores each with No upsampling I think.

OT

I think that Taiko is doing no upsampling, not because their system cannot do it well, but for marketing reasons. It was actually explained by Taiko designer too. They are after end users who own very capable DACs (read: very expensive) which is turn either prefer not to upsample internally or upsample rather well already because of dedicated resources.

 

Personally, I think a double processor, multi-cores system would be extremely good at playback with upsampling or no-upsampling.  

[flkin]

My single Pink Faun 2.16x with 4 extreme OCXO clocks sounds pretty good to me. Can’t imagine what another external box would bring apart from more CPUs, power supplies, connectors, jitter and noise.

 

The new extreme clocks really made a difference - subjectively I’d say something like 30% more pleasure. Not necessary to upsample with the clocks anymore. Everything is so clear and natural already, I’m finding upsampling now only brings a sense of unnatural massaging to the music.

 

One thing about a better clocks, i’ve never found them to have any downsides, only adds to the sound.

 

If you wanted to separate the audio functions, streamer from renderer, the split CPU cores functions in AL has that option. Shocking what details can be extracted with this function but unfortunately the overall presentation becomes too hard and digital sounding for long time listening. 

[cool_chris]

57 minutes ago, flkin said:

My single Pink Faun 2.16x with 4 extreme OCXO clocks sounds pretty good to me. Can’t imagine what another external box would bring apart from more CPUs, power supplies, connectors, jitter and noise

So tell me why the manufacturer used TWO Pinkfauns 2.16x at Munich show ?

I am sorry to tell you but your imagination has nothing to do with what is real.

Also this is the pace were we post What we know .

 

[cool_chris]

1 hour ago, flkin said:

The new extreme clocks really made a difference - subjectively I’d say something like 30% more pleasure. Not necessary to upsample with the clocks anymore. Everything is so clear and natural already, I’m finding upsampling now only brings a sense of unnatural massaging to the music.

This is fact. We know it.

the problem is what we dont know.

Why Taiko extreme is far better sounding than pinkfaun 2.16x ?

If Taiko is using stock clocks ?

There must be something we dont know and Taiko staff knows.

Taiko have tried to add the super moded clocks and it was always worst than with original clock.

I really wanted to know why is that ?

 

[gdpr]

Have you ever performed an inhome comparison test betwwen the Taiko Extreme and the PinkFaun 2.16x?

I have not, and  I am convinced it would cost me several weeks, before I would be able to ‘define’ the difference between them, if at all.

 

I am no saying the Pinkfaun is better than the Taico.

I am extremely happy with the 1-box Pinkfaun server and although still very expensive, it is nowhere in the order of 30k€. And The processor used is AMD 7 Ryzen 2600X processor with 8 cores and 16virtual processors.

My next upgrade will be an update of Audiolinux to be able to dedicate processor to different tasks and applications.

Dirk

[cool_chris]

1 hour ago, ddetaey said:

Have you ever performed an inhome comparison test betwwen the Taiko Extreme and the PinkFaun 2.16x?

No . But i know some who did. They are sure the Taiko is a different animal. Much better in every way. I think base cost of 11K  PF and 20K for Extreme reflects the difference in SQ.

companies value the SQ. If there is no value no one will pay for this 😄.

good question is what about 2 PF  Connected ? This combo being at the price of extreme might be close or better than Taiko . No one tested it .

 

 

 

1 hour ago, ddetaey said:

I have not, and  I am convinced it would cost me several weeks, before I would be able to ‘define’ the difference between them, if at all.

You are asking me  if I had inhome comparison and a line leter „ you are convinced „ there will be no difference without hearing.

I value your opinion but please do not deny youself 

 

[cool_chris]

5 hours ago, lmitche said:

Get the Taiko Audio SGM Extreme. It is the best of the best. Did you know the Totaldac D1-twelve mk2 is one of the DACs used when testing the SQ of the Extreme?

 

I know a lot about the Extreme. I am impressed with the build and science behind that.

seriously considering to get it. Very expensive but probably the best solution.

I am just unsure if stripped Windows will work as good as the new installed device years later.

I have very bad memories with W 12 server. It sounds great once optimised.

afer a month or two it is garbage . SQ is very compromised.

you have to get back and install everything including the system , players and optimization.

i think Euphony could be better here than Windows.

[lmitche]

1 hour ago, cool_chris said:

I am just unsure if stripped Windows will work as good as the new installed device years later.

It would be very surprising if Windows automatic updates are not disabled on the Extreme. Also, the Taiko people will bend over backwards to keep your system up to their latest release. Check for yourself, but I wouldn't have any worries about Windows or Taiko service ethic.

[flkin]

4 hours ago, cool_chris said:

So tell me why the manufacturer used TWO Pinkfauns 2.16x at Munich show ?

I am sorry to tell you but your imagination has nothing to do with what is real.

Also this is the pace were we post What we know .

 

 

I have no idea why Jord used two 2.16x machines although he did tell me that's the next upgrade. He might be right but my pocket has difficulty believing it 😅

 

I repeat, I really can't imagine how much better things can get with another set of components massaging the signal. 😄😂

 

 

[bobfa]

Has anyone gone back and re-evaluated the USB chain?  

 

I decided on something rather simple an LPS-1.2 powering an ISORegen in 2017.  I have been "afraid" to take it out and listen without it,  Silly!   Today I am going to try the USBPCB to the KII Control rather than the second cable.  Just for fun.  

 

I went back in the thread and looked around some more, but did not find any evidence of a newer way of doing things.  Say just using a Pink Faun or Wolf Audio USB Card and a single cable?

 

This post is not meant to open up another workload for me or others, just trying to keep my ear to the ground.