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

[lmitche]

20 minutes ago, seeteeyou said:

Thread gone quiet lately?

CTU,

 

We are busy listening to music on our absolutely awesome NUC/Audiolinux systems. The recent software upgrades are fantastic! SQ exceeds my best expectations.

 

Larry

[austinpop]

7 hours ago, seeteeyou said:

Thread gone quiet lately?

 

6 hours ago, lmitche said:

We are busy listening to music on our absolutely awesome NUC/Audiolinux systems. The recent software upgrades are fantastic! SQ exceeds my best expectations.

 

While I have a different configuration, I agree with Larry - we're at a point now where most of us are happy with the excellent SQ we are enjoying. Here is what my current system looks like:

 

 

I am alternating between AL and Euphony as the OS for the Server and the Endpoint. With the recent changes, AL has almost, but not quite, caught up in SQ with Euphony. I own both, so continue to monitor improvements and can easily switch back and forth as needed.

 

One very interesting finding, which I will expand on in an upcoming review, is that with the improvements in my network chain with the bridged JCAT Net Femto card and the SOtM switch, the SQ through the Ethernet interface of the DAC is now very close to the SQ through the USB input. This is true across the Ethernet DACs  at my disposal, the Ayre QX-8, the QX-5, and my current DAC under review. The USB path still wins out, but by a smaller margin than ever before. Of course, with USB you can also play with alternated endpoints like Squeezelite, buffers, etc. Still, this suggests that there may be an alternate path to great SQ using just reclocking switches and Ethernet DACs. This has the benefit of being a low-spaghetti solution.

 

I am currently trying to see if I can systematically force my entire chain to 100Mbps. Like many here, I am finding 100Mbps to sound better than 1Gbps.

 

Eric @limniscate and I rediscovered the advantage of 100Mbps in a recent listening session on his system, which he recently upgraded with the fabulous Maggie 20.7's. In addition to enjoying the music, we wanted to test the effect of the SOtM switch powered by an SR7 DR rail, on the Ethernet input of the DAC under review. We set up the USB chain as the baseline:

• Wall (router) > Zenith SE > tX-USBultra (SR7 DR) > DAC

Note - we used SOtM dCBL-Cat7 ethernet cables and Lush USB cables for all experiments.

 

We next tried the SOtM switch upstream of the SE:

• Wall (router) > SOtM switch (SR7 DR) >  Zenith SE > tX-USBultra (SR7 DR) > DAC

Oddly enough, this sounded worse. Upon investigation, we found from the lights on the SE's Ethernet port that in the baseline case, when connecting directly to his router, his SE was negotiating down to a rate of 100Mbps. With the switch in the path, the switch was auto-negotiating 1Gbps both to the router, and to the SE. Net net (pun intended!) - the SE in the baseline case was running at 100Mbps and sounding better.

 

Next, we tried a direct path from the SE to the DAC via Ethernet:

• Wall (router) > Zenith SE (bridged) > (via Ethernet) > DAC

This sounded significantly worse than the USB path. Soundstage depth was lacking, and the overall presentation was flat and boring. Since the DAC only supports 100Mbps, in this case both ports on the SE were at 100Mbps.

 

Then we added the SOtM switch between the SE and the DAC:

• Wall (router) > Zenith SE > SOtM switch (SR7 DR) > DAC

In this case, the speed negotiated by the ports were:

• SE LAN: 100Mbps (to router)
• SE streamer: 1Gbps (to switch)
• Ethernet DAC: 100Mbps (to switch)

This was by far the most dramatic improvement I've ever heard from an Ethernet switch. Comparing this path to the baseline USB path, the USB path still won. The gap was small running Roon as the endpoint, but widened again when using squeezelite (the so-called experimental option on the Zenith SE).

 

As I try to force 100Mbps end to end, I find myself wishing for a 100Mbps/1Gbps config option on the SOtM switch. While a case can be made that the speed should properly be configured on the nodes, many OSes like InnuOS and Euphony don't allow any way to override the auto-negotiate setting of the Ethernet interface. Even on AL it is a bit fiddly, and best left to the Linux experts.

 

As far as what is next on my system - not much! I am finally very happy with my digital path upstream of my DAC. The next big thing for me is power supply improvements. I am eagerly awaiting (and awaiting, and awaiting) my 3-rail SR7 DRXL, which will allow me to run my NUC, tX-USBultra, and SOtM switch on SR7 rails. That should be it for me until I can try the Uptone EtherRegen switch against the SOtM and TLS switches.

 

I so want to be DONE. I say that, but knowing myself, tweaking will likely continue!

[Ralf11]

you might add that one to your compilation

 

also: the blue arrows point down to your AC power - maybe up would be better

 

& would 'dedicated AC line' be better wording than 'dedicated circuit' - esp. for non-native English speakers?

[austinpop]

2 hours ago, Ralf11 said:

you might add that one to your compilation

 

also: the blue arrows point down to your AC power - maybe up would be better

 

& would 'dedicated AC line' be better wording than 'dedicated circuit' - esp. for non-native English speakers?

 

Good suggestions. I've updated the post.

[elan120]

After getting the NUC back from SOtM clock mod along with upgrading to the latest AL version 1.2.0, the motivation to do more tweak is for sure lowered by a large margin.  I managed to change couple more items in my NUC, with more than two weeks of continuous burn-in, they helped elevate the SQ up by a good notch.

 

First mod is more universal by adding Evox cap to NUC.  This mod improved the overall dynamic, having similar effect  like other Evox upgrades, and it takes at least a week to get the cap settle in.

 

The second mod is more suited to my modified NUC by replacing the stock DC cable to sCLK-EX board with Audio Sensibility's silver Neotech cable, terminated in JSSG360 configuration.  This change lowered the noise floor quite nicely, and this mod can be implemented for other devices having internal DC cables such as tX-USBUltra.

 

There are few more things I plan to experiment with my front-end section still, but with these changes, I seems to procrastinating quite a bit by just enjoying the music lately.

[austinpop]

8 minutes ago, elan120 said:

After getting the NUC back from SOtM clock mod along with upgrading to the latest AL version 1.2.0, the motivation to do more tweak is for sure lowered by a large margin.  I managed to change couple more items in my NUC, with more than two weeks of continuous burn-in, they helped elevate the SQ up by a good notch.

 

First mod is more universal by adding Evox cap to NUC.  This mod improved the overall dynamic, having similar effect  like other Evox upgrades, and it takes at least a week to get the cap settle in.

 

The second mod is more suited to my modified NUC by replacing the stock DC cable to sCLK-EX board with Audio Sensibility's silver Neotech cable, terminated in JSSG360 configuration.  This change lowered the noise floor quite nicely, and this mod can be implemented for other devices having internal DC cables such as tX-USBUltra.

 

There are few more things I plan to experiment with my front-end section still, but with these changes, I seems to procrastinating quite a bit by just enjoying the music lately.

 

Thanks for that update Kevin!

 

You actually answered a question I meant to ask at the end of my last post but forgot:

• are you folks who've been making similar enhancements also finding yourself on a similar plateau?

[mourip]

11 hours ago, austinpop said:

 

Thanks for that update Kevin!

 

You actually answered a question I meant to ask at the end of my last post but forgot:

• are you folks who've been making similar enhancements also finding yourself on a similar plateau?

 

I too was wondering why is was so quiet here 🙂

 

I would have to say that I have found the much more flat part of the curve. Increasingly the changes I make may produce a "different" sound but not necessarily a step up. Now the challenge is finding well recorded and personally enjoyable music. Thanks to Qobuz that is not too hard these days.

 

I have a JCAT Femto ethernet card on the way. Hope springs eternal...

[Dutch]

58 minutes ago, mourip said:

I have a JCAT Femto ethernet card on the way. Hope springs eternal...

 

 Me too!  Unfortunately it was/is delayed but I did receive a UPS notice last week (though it’s not en route yet) so it must be coming soon to us!

[Ricardo007]

On 5/20/2017 at 11:37 AM, amir57bs said:

 

Usb cable in my idea is very very important .

overally if you like emotional sound (true tone plus good micro dynamic) i think combination of gordon rankin codes plus purist cable is very good.

i do not like ultra smooth and clean sound of many systems , i prefer emotional sound .

i think ethernet renderer may give you ultra clean ultra smooth ultra low noise sound but i prefer intimate sound of pcie usb card with AS gordon rankin codes.

the best way to get emotional sound is to keep the way minimal.

adding reclockers or ethernet devices like microrendu add complexity of streaming data transfer and it will degrade intimacy of sound.

What rare "AS Gordon Rankin codes"?

[dgarretson]

I have a few hundred hours on a JCAT Femto  NIC card powered by an ebayed Zerozone 2A LPS.  In agreement with Austinpop, the JCAT is a game-changer--- compared to a standard  EN port in my QNAP TVS-882 NAS.  The $130 Zerozone was a gamble, but seems to have paid off.  A second one made a major improvement relative to the wall wart I had been using to power a SOtM-modified D-Link hub.  Everywhere else I'm SR-7, but the Zerozone seems to be a reasonable compromise.  I'll probably get a third one to power my Verizon router. 

[One and a half]

On 4/23/2019 at 10:37 PM, Dutch said:

 

 Me too!  Unfortunately it was/is delayed but I did receive a UPS notice last week (though it’s not en route yet) so it must be coming soon to us!

Don’t worry , the JCAT won’t disappoint. For best results, power the board with a separate 5V , 2.5A psu , sinks the noise floor down a few notches.

Must try the speed limit to 100Mb/s & see what happens.

[kyrill]

On 4/23/2019 at 12:39 AM, elan120 said:

After getting the NUC back from SOtM clock mod along with upgrading to the latest AL version 1.2.0, the motivation to do more tweak is for sure lowered by a large margin.  I managed to change couple more items in my NUC, with more than two weeks of continuous burn-in, they helped elevate the SQ up by a good notch.

 

First mod is more universal by adding Evox cap to NUC.  This mod improved the overall dynamic, having similar effect  like other Evox upgrades, and it takes at least a week to get the cap settle in.

 

The second mod is more suited to my modified NUC by replacing the stock DC cable to sCLK-EX board with Audio Sensibility's silver Neotech cable, terminated in JSSG360 configuration.  This change lowered the noise floor quite nicely, and this mod can be implemented for other devices having internal DC cables such as tX-USBUltra.

 

There are few more things I plan to experiment with my front-end section still, but with these changes, I seems to procrastinating quite a bit by just enjoying the music lately.

There are a lot of experienced DIY hobbyists  reading  this topic. Which knowledge can be shared so more readers can replicate the Evox mod by doing it themselves? IS it somehow related to the power?

[Nenon]

Regarding the JCAT Femto NET card...

 

I am sure it would make a difference when streaming Tidal, Qobuz, or even files over the network. No doubt about that!

 

But let's assume I am playing local files only.

I have noticed that my NUC is very sensitive on noise getting through my network cable. If I play a (local) file and disconnect the network cable while playing on my NUC, I can hear an improvement a few seconds later. I plug in the network cable, and the sound quality gets worse. It is quite audible. It was also shocking the first time I heard it. I can see how the JCAT Femto NET card would make a difference in this case. You need to control your software somehow, and typically you do this over the network. In that case even playing local files the JCAT Femto NET card would be an improvement. 

 

However, this is when it gets interesting. I have another computer I am playing around with. On this one I cannot hear any audible difference when playing local files with the network cable plugged or unplugged. Looking more into the motherboard architecture and the NIC components, I do see some isolators implemented on the NIC. So perhaps those isolators are good enough to eliminate most of the noise coming over the network. 

Can anyone think of a way how the JCAT Femto NET card would be an improvement for locally played music on this computer? 

[RickyV]

On 4/22/2019 at 7:27 PM, austinpop said:

I am alternating between AL and Euphony as the OS for the Server and the Endpoint. With the recent changes, AL has almost, but not quite, caught up in SQ with Euphony. I own both, so continue to monitor improvements and can easily switch back and forth as needed.

So Euphony on server and endpoint still got better sq then a single Euphony server alone? Have you tried a Euphony server and a al/Squeezelite endpoint config, so you still play with the buffers or would the degrade the Euphony/euphony sq?

[austinpop]

2 hours ago, Nenon said:

Regarding the JCAT Femto NET card...

 

I am sure it would make a difference when streaming Tidal, Qobuz, or even files over the network. No doubt about that!

 

But let's assume I am playing local files only.

I have noticed that my NUC is very sensitive on noise getting through my network cable. If I play a (local) file and disconnect the network cable while playing on my NUC, I can hear an improvement a few seconds later. I plug in the network cable, and the sound quality gets worse. It is quite audible. It was also shocking the first time I heard it. I can see how the JCAT Femto NET card would make a difference in this case. You need to control your software somehow, and typically you do this over the network. In that case even playing local files the JCAT Femto NET card would be an improvement. 

 

However, this is when it gets interesting. I have another computer I am playing around with. On this one I cannot hear any audible difference when playing local files with the network cable plugged or unplugged. Looking more into the motherboard architecture and the NIC components, I do see some isolators implemented on the NIC. So perhaps those isolators are good enough to eliminate most of the noise coming over the network. 

Can anyone think of a way how the JCAT Femto NET card would be an improvement for locally played music on this computer? 

 

Very interesting findings about the SQ variations simply due to the "presence" of a network connection. I have found the same, and it varies by machine, and what is upstream of the box, in terms of switches, the network speed (100 vs. 1000 Mbps), etc.

 

It appears the new computer you're playing with does not degrade SQ with a network cable attached. In this scenario, if you're not planning to actually stream data, and your only use for the network is for control purposes, then yeah I'd think the JCAT card is overkill.

 

But like most of these things, actual experience may be different.

[austinpop]

31 minutes ago, RickyV said:

Have you tried a Euphony server and a al/Squeezelite endpoint config?

 

To be clear - I'm a Roon user. So all my comments have been about Roon Core on the server, and Roon Bridge/SL/SL+large buffers on the endpoint.

 

However, to answer your question: yes, I've tried all the combinations of AL and Euphony, both on server and endpoint. The latest AL is now very close to Euphony SQ, but Euphony still sounds a little better to me.

 

Quote

so you still play with the buffers or would the degrade the Euphony/euphony sq?

 

Squeezelite with large buffers is possible on both AL and Euphony. It makes the same improvement on Euphony as it does on AL.

 

Quote

So Euphony on server and endpoint still got better sq then a single Euphony server alone?

 

Standalone server is a new use case for me, and I''ve just started to explore that, comparing Euphony/Roon and Euphony/Stylus. Too early to say much other than Stylus sounds very, very good. I've asked Euphony if they can make Stylus an endpoint to Roon, and they said they'll consider if it's feasible. Here's hoping!

[RickyV]

7 minutes ago, austinpop said:

To be clear - I'm a Roon user. So all my comments have been about Roon Core on the server, and Roon Bridge/SL/SL+large buffers on the endpoint.

Sorry forgot about that.

[Nenon]

18 minutes ago, austinpop said:

Very interesting findings about the SQ variations simply due to the "presence" of a network connection. I have found the same, and it varies by machine, and what is upstream of the box, in terms of switches, the network speed (100 vs. 1000 Mbps), etc.

 

Actually I do hear some difference on the other server as well. Double checked now. It's just not as pronounced as it is on my NUC.

The interesting thing is that my streamer connects to a 100 Mbps fiber media convertor (tweaked by me with some MUSE capacitors and feb by LPS2.1). I never played with hardcoding the speed/duplex. But I also always found the sound quality better when using the media convertor. I always thought it sounded better because of the further isolation. But now I am thinking it has double effect as it also forces my streamer to negotiate at 100 Mbps. 

 

18 minutes ago, austinpop said:

Standalone server is a new use case for me, and I''ve just started to explore that, comparing Euphony/Roon and Euphony/Stylus. Too early to say much other than Stylus sounds very, very good. 

 

I have compared AudioLinux / Roon vs. Euphony / Stylus on a standalone server in my system. I prefered Euphony / Stylus. But again, I did not compared apples to apples as the software and the OS were different. 

[mourip]

11 hours ago, One and a half said:

Don’t worry , the JCAT won’t disappoint. For best results, power the board with a separate 5V , 2.5A psu , sinks the noise floor down a few notches.

Must try the speed limit to 100Mb/s & see what happens.

 

I have my Roon server working in bridged mode at 1G but am willing to try at 100mhz to see if it improves things. Does anyone know if AL will auto-negotiate on the endpoint PC if I change the server setting?

[austinpop]

2 minutes ago, mourip said:

 

I have my Roon server working in bridged mode at 1G but am willing to try at 100mhz to see if it improves things.

 

Please try it and tell us what you find.

 

2 minutes ago, mourip said:

Does anyone know if AL will auto-negotiate on the endpoint PC if I change the server setting?

 

If the endpoint is connected directly to the server, with no intervening switch, then yes, it will auto-negotiate.

[numlog]

In my PC the SATA card and MoBo both use 25MHz clocks which are currently some cheap general purpose XOs used to test initially if replacing the crystals with an external clock would actually work. They will be getting replaced now and the big thing on my mind is if I could get away with running the MoBo/Chipset and SATA PCIe card off the same clock, any thoughts on that?

 

One drawback of using 1 XO is a longer wires ( ~2cm to ~15cm) from the XOs output to get to both inputs.

The SATA card and Chipset/MoBo clocks would be synchronised, like the internal SATA ports would be... a good, indifferent or bad thing? not sure 

[ray-dude]

3 hours ago, austinpop said:

 

Please try it and tell us what you find.

 

 

 

Interesting, using 

ethtool eno1

 I just checked the ethernet speed on my 7i7 NUC endpoint (running AudioLinux) and I'm getting 10Mb/sec(!)

 

That may explain why I've had some skipping issues with very high res files

 

Time to figure out how to set or what is causing the low negotiated speed

 

 

[Bricki]

3 hours ago, austinpop said:

 

Please try it and tell us what you find.

 

 

If the endpoint is connected directly to the server, with no intervening switch, then yes, it will auto-negotiate.

G'day Rajiv

 

How can we reduce our network speeds to 100mbps on Audiolinux?

Sorry for the dim question 🙄. I'm not a Linux native 😁

 

I have my server and endpoint bridged.

 

Thanks

Brendan

 

[Dev]

With Ethernet interfaces, you can use ethtool utility in Linux. However, AL doesn't have this package by default. So you have to download it using pacman or yaourt

For example to set link speed to 100mbps on eno1 interface,

 

ethtool -s eno1 speed 100 duplex full autoneg off

 

[Bricki]

25 minutes ago, Dev said:

With Ethernet interfaces, you can use ethtool utility in Linux. However, AL doesn't have this package by default. So you have to download it using pacman or yaourt

For example to set link speed to 100mbps on eno1 interface,

 

ethtool -s eno1 speed 100 duplex full autoneg off

 

Ok thanks for this 👍