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

[shahed99]

Is it better to have a 16GB ram than 8GB for AL NUC? This up squared fanless system seems like a screaming deal for NUC use.

 

https://up-shop.org/home/258-up-squared-edge-system-powered-by-intel-x7-e3950-soc-8gb-ram-64gb-emmc.html

 

It comes with:

* Intel ATOM Quad Core up to 2.0GHz x7-E3950

* 8 GB RAM

* 2x GB Ethernet (Is that the ideal for direct connection to upstream server?)

*64GB eMMC Storage (Is there any point of eMMC storage if someone is booting from AudioLinux on USB?)

 

The price seems very reasonable at $399. Anyone tried this as NUC with AL?

[shahed99]

Also, another question/observation regarding AL on NUC:

 

The recent experiments suggest that AL on NUC as a streamer/endpoint is a screaming value beating out fully tricked out SoTM tri-fecta and similar devices. However, is there any intrinsic benefit of using a dedicated NUC + AL (or any other computer for that matter) for streaming/Roon endpoint instead of relying on a fully tricked out one single server (linear PSU + sCLK-EX modified mobo+USB PCIE) direct connection (with or without SoTM USBUltra) to DAC usb input?

 

I understand that a big benefit of isolated NUC is that your upstream computer can be noisey or less than perfect. When someone is building a dedicated fanless music server with proper PSU/clocking/audio grade USB PCIE - is there any benefit of using a streaming NUC+AL?

 

It'll also be interesting to see in a single server situation using AudioLinux and compare to Windows Server 2016 + Audiophile Optimizer.

[shahed99]

15 minutes ago, Dev said:

 

Yes, the NUC8i7BEH model.

 

I don't think lps-1.2 can power that model. For now I will use the supplied smps but need to figure out a good lps. Will use this on the server end and experiment with Roon/HQP/LMS, whichever sounds better in my setup and report back. For my use case, the NUC needs to do at least dsd256 and be stable and run cool. Will see.

Curious if you have a fanless case for the NUC8. If not, fan can have some negative impact on sound quality by creating ripples (even when the fan is not audible). However, the point is moot if you are using another NUC to stream.

[shahed99]

I recently won an ebay auction on NUC7i7DNBE in a Akasa case with 16GB ram for a great price. What a timely thing to do!

 

@austinpop As always, thank you for sharing your impression with us. Is it possible to discuss the best way to connect upstream server and NUC/AL endpoint? For example, @lmitche found wifi to have the best sound quality? Others have mentioned a OCXO switch by TLS. Any comparison of such solutions with AL/NUC as the endpoint will be very helpful.

 

I'll also be curious to hear NUC server and AL/NUC endpoint comparison with sCLK-EX'd server and AL/NUC endpoint. 

[shahed99]

21 minutes ago, BigAlMc said:

 

 

Endpoint NUC

 

Average to good CPU like the Celeron in NUC7CJYH - better is fine but will cost more

4GB to 8GB RAM - more is fine but will increase power usage

No SSD as will add noise 

More Cores and more CPU cache should sound better

Low power usage is better especially if you want to use an LPS1.2 on 12v or similar

 

 

 

I believe it has been found that i7 NUC also sounds better as an endpoint

[shahed99]

@austinpop Did you get a chance to listen to your clock modified NUC + USBUltra + SoTM Switch without the master clock input from Mutec? I was wondering, if I should send my NUC and USBUltra to SoTM, should I get the master clock mod. Given the price of a Mutec and good digital cable is $5k, I wonder how significant it is to have it in the chain. 

[shahed99]

What a wonderful report by @romaz! 

 

I recently got a SoTM sNH-10G switch with sCLK-EX, master clock input (although not using master clock as I don't have one) and updated EVOX cap. Powering the switch with Upton LPS 1.2, I noticed a satisfying uptick in sound quality right away. Not only Tidal streaming sounds better but also local files is seeing an improvement in sound quality. I'm using AL/NUC i7 endpoint, so this is perhaps expected.

 

However, the point of the post is how we all hear things differently. I bought a new MikroTik router because it has 10G SFP+ fiber port. I wanted to connect the SoTM switch using its SFP port to the SFP+ of my new router. Previously my SoTM switch was connected using a Blue Jeans 6A patch cable about 5' long. Now, with optical connection to replace the 6A patch cable, I heard less noise and a calmer representation of music. There was a slight reduction of high frequency noise.

 

I'm sure, if I used audiophile grade LAN cable from SoTM or PAD, it would sound better than the optical connection. But, given I only spent $30 (optical cable + module), I'm very happy with the improvement.

 

Need to use a better power supply to power my SoTM switch. As reported by romaz, the PSU needs more headroom to get the best performance out of the switch.

[shahed99]

@valveboy I've been in this hobby too long and experimented with too many things to know that the thread you posted is not very useful to me. I've a supra ethernet cable that sounds objectively different (subjectively better) than my Blue Jean's patch cable. Please dont let yet another cable debate to derail this thread. 

[shahed99]

3 hours ago, auricgoldfinger said:

 

Audio Sensibility has their own wire, described below.

 

 

 

@auricgoldfinger Did you order the wire from Audio Sensibility and did the mod yourself or had to send the txUSB-Ultra to them to have it done? I'm planning to send my Ultra to SoTM to do the evox cap upgrade. I wish they had the Audio Sensibility wire to do the upgrade.

 

I've used Audio Sensibility interconnect in the past. They are great cables for a reasonable price.

[shahed99]

Note: Somewhat long post. TLDR provided at the bottom of the post.

 

Something "massive" has happened last night that left me scratching my head. Before I proceed further, just want to clarify that it is heard and observed in my system by my ears. So, theres' no objective validity except my subjective perception.

 

As I've mentioned in my earlier post, I recently bought a fully tricked out SoTM sNH-10G switch connected to a Mikrotik router via SFP port. The sound improvement was very encouraging. Inspired by the reports from Rajiv and Roy, I decided to send my USBUltra and NUC (NUC7i7DNHE) to SoTM for clock and Evox cap upgrade. I disconnected my NUC to pack it up last night. But out of curiosity, I still kept the USBUltra in the chain. Now, my RoonServer is directly connected to the DAC via USBUltra. I was expecting a downgrade in sound. But, the sound actually improved. In a small but very satisfying way...

 

Now, on to the details. I'll have to provide some background information to give proper context to this finding. I bought a pair of Wilson Sabrina after I moved to my new apartment last Fall. I didn't have enough room treatment that time and was struggling with the Sabrina. It is a fantastic speaker but it was hard to listen to sub par recording  as those would sound bright. So, I sold it few months ago and ended up gettin a pair of Harbeth. In between the speaker switch, I got the AL/NUC combo in a Akasa fanless case to use it as Roon endpoint.

 

Harbeth is a lovely speaker and made the sub par recording sound very listenable without losing any details. However, it wasn't still taking me all the way. There was still be some edginess and overall leanness to the sound that would bother me. I would love listening to music for an hour or two. Longer than that, I'll feel overwhelmed with slight headache. Even though, I now have extensive room treatment, I thought my room is the one to blame.

 

Last night, after moving to my single server setup (taking the NUC off the chain), the leanness of sound has gone. There is slight reduction of edginess. I might have lost a tiny amount of details and resolution. But the sound is fuller, little thicker and weighty in a pleasant way.

 

I built my music server almost 5 years ago with i7 4770T, paul pang provided TCXO mobo, TCXO SSD, RAM and SoTM USB PCIE card. Powered by Teradak ATX PSU and various other linear and battery PSU for SSD and USB PCIE. Although decent, but nothing to write home about. This server has WIndows Server 2016 installed, optimized with Audiophile Optimizer.

 

I've two possible explanation for the sound improvement:

 

1) WS 2016 + AO: I think I prefer the sound of Windows. My ears are particularly sensitive to thin/bright sound. In that respect, Windows does a better job than Linux IMO.

 

2) ASIO driver on Windows: I'm able to use HQPlayer upsampling from PCM to DSD in native mode. On Linux platform, only DoP was supported and upsampled DSD didn't sound good; hence I never ended up upsampling DSD while using AL/NUC. However, I should mention that comparing playing PCM only via AL/NUC (either RoonBridge or SqueezeLite) vs single serer setup, the later came ahead with smoother, weighty and fuller sound.

 

3) A third possibility (albeit I'm not convinced): I was using LPS 1.2 to power the SoTM switch in sPS-500 to power the AL/NUC. In the single server mode, I used the sPS-500 to power the SoTM switch and USB PCIE card. sPS-500 can provide 5A current. So, it can give more head room even when powering both the SoTM switch and USB PCIE compare to LPS 1.2  (1.1A) powering the switch. I'm sure the headroom in powering the switch has helped the single server system a bit. But, I'm not sure that is all there is. I heard the improvement in sound quality through my single server even when playing music stored locally.

 

I waited a full day before reporting to make sure what I hear is really there. I've listened to music almost 8 hours today without any fatigue. At this point, I don't think I'm going to send the NUC to SoTM for clock upgrade. I've been hearing that WS 2019 sounds even better than WS2016 and AO 3.0 is going to be out soon. With all that happening, I' think I'm going to sell my NUC and invest fully in single server system. I hear that Euphony sounds fuller and more dynamic. However, given I'm already getting the fuller/weighty sound from my WS2016 music server, I'm not very keen to go on that route.

 

I ordered a Reference PSU from Adrian (of TLS) to use with my NUC. I've heard through the grapevine that new version of Reference 1 is sounding as good as PH SR7. But given this new revelation, I don't know what I'm going to use the new PSU for. I'd need a good quality ATX PSU instead for my Roon server. If the upcoming SoTM motherboard has DC input, then it will be great.

 

I honestly don't think my finding invalidates the superb sound quality people getting from AL/NUC endpoint. It's just that in my system, I happen to prefer my single music server setup. I would mention one important thing though. People who has used Windows Server + AO in their setup and like the windows sound in general, I'd advise to do careful comparison before jumping to a linux based server (be that NUC or custom build). There is a fundamental sound characteristics difference between Windows and Linux OS.

 

TLDR:

Moved from i7 AL/NUC (powered by sPS-500) endpoint to single server system (WS 2016+AO). Noticed that sound is now less lean, fuller, thicker and weighty. I find it more musical and great for fatigue free listening. Contemplating to sell the NUC.

[shahed99]

15 hours ago, romaz said:

I'll venture a guess.  Your optimized server is better than your unoptimized NUC.

That seems to be the case. When You/Rajiv and others reported the finding in regards to NUC, it was mind blowing. Mostly because, how can an-off-the shelf hardwire can sound better than purpose built music server. The fact that NUC without any modification holds up so well to a purpose built music server is astounding.

 

15 hours ago, romaz said:

Put something like an SR7 or the TLS Reference 1 to the NUC and you may change your mind about the NUC sounding thin.

I wish I had spare PSU lying around to test these scenarios. It would have been very educational. I've no doubt the NUC would scale really well with great power supplies like that. 

 

15 hours ago, romaz said:

Also, if you're using an independently powered SOtM PCIe USB card in your server, this is obviously giving your server an edge over your unmodified NUC.  Try connecting your server to your DAC using one of standard integrated USB ports on your motherboard and the delta that you hear might give you some idea of how much better your NUC could sound once you send it to SOtM for clock replacement.

I've used my motherboards standard integrated USB port and the sound is frankly unlistenable. Goes to show, how good the USB output of the NUC is. However, not sure if the clock improvement is the only thing to close the delta. As the SoTM USB PCI-E has other things like filters and special voltage regulators. 

 

15 hours ago, romaz said:

As for WS2016 + AO sounding thicker, fuller and weightier than AL, the easiest way to find out in your system is just to boot your server into AL using a USB stick and compare.  They both can be tuned to sound fuller or leaner.  Regarding Euphony vs AL, yes, to my ears, Euphony sounds fuller and weightier and so it will cost you nothing to give Euphony a try but ultimately, if you're already happy with what you're getting from just your server, perhaps it's time to sell the NUC and stay with what you already have.  Finding the perfect balance is a tricky thing and what works for one person may not work for another but once you find that perfect balance, it's golden.

If I end up keeping my NUC, I will either use Windows Server on it (if possible) or try the Euphony. 

 

You and Rajiv reported that adding a NUC (with no clock modification) as a Roon endpoint to a high quality server like Innuos Zenith was an improvement in sound quality compare to using the Innuos alone. My custom built server I'm sure is not as good as Innuos. But still, it sounds better on its own comparing to AL/NUC endpoint. That has left my head scratching.

 

Many of you got great result with the unmodified NUC using it as Roon endpoint compare to stand alone server. That is not the case in my situation. That is why I'm in a dilemma doing the clock upgrade and better PSU no doubt will yield a better sound quality but whether the ROI would be worth it. I think I might try installing Windows Server on the NUC and see the result.

 

As a side question: May told me that it is possible to install sCLK-EX inside the NUC but would have to remove the 2 USB connection from the front of the case. Do you think there's added benefit of having a dedicated sCLK-EX inside rather than through USB Ultra?

[shahed99]

I decreased my ethernet speed to 100Mbps on Windows Server 2016. I noticed a calmer and more relax presentation. Very nice!

[shahed99]

Can anyone recommend which Optane drive to use with my NUC 7i7DNHE? I was looking into this one on amazon and the compatibility tool there says it's not sure if Intel NUC is compatible with that particular Optane. 

[shahed99]

@Nenon Thank you for facilitating purchase of Apacer RAM. I am eagerly waiting for mine.

 

I became a RAM believer when I build my dedicated music server using all Paul Pang components in early 2015. I recall when I was waiting for Paul Pangs special RAM (non ECC) to arrive, I was using regular consumer grade RAM. Once I replaced them with PPA RAM, the improvement in sound quality was immediate, noticeable and quite significant. I still have those although they are DDR3 RAM. I'm curious what Paul Pang has up its sleeve lately.

[shahed99]

1 hour ago, taipan254 said:

Apart from aesthetics, is there any downside in using a regular ATX case with passive heatsinks, as opposed to heatsinks that are built into the chassis? Something like this, for example, could, in theory, cool a higher TDP processor (especially if underclocked / "tuned" to a certain frequency) while allowing you to build in an ATX form factor. It doesn't look as cool as a Streacom case, but it would get the job done. Why limit yourself to a smaller form factor if you don't have to?

 

There are primarily two reasons for using passive cooling: i) noise and ii) resonance/vibration induced by moving parts. These days there are silent fans (Magnetic Levitation Cooling Fan) available that are practically inaudible. However, the fact still remains that those are moving parts that causes vibration/resonance on the motherboard, which is not good for sound.

 

The only solution to keep using a passive case (HDPLEX/Streacom) is to use a CPU with TDP 65W. However, as reported by both @romaz and @austinpop, beefy CPU with high core counts, large L3 cache and higher frequency sounds beefier and dynamic. But those often has greater than 95W TDP. 

 

The challenge is how to passively cool a CPU that has TDP >= 95W. @seeteeyou has posted some links about upcoming passive chassis that might be capable of supporting TDP 250W looks promising. But one solution that I see that is rarely mentioned is water cooling.

 

I recently built a music server using Ryzen 3950X (16C/32T/64MB L3) with TDP 105W. 7nm CPU are more power efficient but generates slightly more heat than 14nm chip. My Streacom FC10 (even with extended heat pipes and kryonaut paste) wasn't enough to cool the beast. Hence, I used a Corsair AIO cooler, which breaks no sweat keeping the CPU temperature under 35C even when running my server nonstop for days. The radiator fans are placed outside of the chassis to reduce vibration and I can't hear them at all sitting 7' away from my speaker. However, given this is a AIO - there's a pump sitting on top of the CPU to move the water through the pipe. It's not audible but still is a moving part. So, not ideal. However, if you can spend money $350-$700, there are custom water cooling solution where radiator/fans/pump everything is external to the chassis and with long tubes those can be further away from your listening position and inaudible. There are no CPU on earth that can't be adequately cooled with custom water cooler solution with no moving parts inside the chassis. 

 

Even without any mobo clock upgrade like my previous build of i7-4770T and linear ATX PSU, this new server with off the shelf switching PSU sounds like a component upgrade over the old ones. It's massively dynamic sounding. While the sound has some high frequency harshness due to using switching PSU, I suspect it'd be gone when I get a linear PSU.

 

[shahed99]

25 minutes ago, Iving said:

 

 

 

- Max CPU Cooler Height: 135mm for e.g. NOFAN CR-80EH Copper IcePipe as well as air, water

 

 

https://streacom.com/products/f12c-chassis/

https://streacom.com/downloads/ug/f12c_150407.pdf

 

 

Still just about available.

 

I got excited when I first saw this case. However, this is not meant to be a passive case. Problem with using after market passive cooler like NOFAN is that either you'd need to install an intake and exhaust chassis fan to dissipate the heat off the passive cooler as there's no heat pipe that is touching the chassis surface to dissipate the heat to outside environment or you need to have the chassis open to have free airflow from outside. But I suspect, in either cases you'd not get TDP 80W of cooling with CR-80EH, more like 65W that precludes the best in class CPU that starts with TDP 95W.

[shahed99]

1 minute ago, taipan254 said:

 

Just a point of clarification - I asked Roy why he doesn't use a regular CPU case with a passive heatsink. One can achieve

the same, well-known benefits you describe above, @shahed99, including 120W TDP compatibility, and use a full ATX motherboard. As you know, passive heatsinks do not have any fans / moving parts / noisy motors. 

 

I bring this up because Roy was lamenting using uATX boards simply for case and cooling compatibility. 

 

There are no proven, ready to purchase passive heatsink available in the market capable of cooling TDP > 100W CPU effectively, at least I'm not aware of any. The Noctual one you've linked, I've been keeping an eye on them for a while. So far, it's just a concept with no indication when or if at all if will be released. I vaguely recall another concept passive heatsink that were suppose to be capable of cooing ~100W CPU but the manufacturer had to abandon the idea for whatever reason.

[shahed99]

14 minutes ago, Iving said:

Looked it up in case of interest: "The CR-95C can continuously and silently cool any processor with a TDP specification of 95 watts and under."

That could be a very effective solution esp for someone who doesn't want/need a streacom/hdplex type passive and make it work with regular chassis. But might need to make sure to not stress the system/CPU and have adequate ventilation and ambient temperature if powering a CPU with TDP of 95W or greater.

 

It's been a while since I watched this youtube video to remember the details where the reviewer tried to power a Ryzen 3900X with CR-95C. As far as I recall, it was useable but the temp was quite a bit on the higher side even when the system wasn't on for that long.  

 

[shahed99]

Romaz has mentioned in his last post that powering the CPU/EPS through HDPlex 800W DC-ATX sounds better than directly powering by a 12V rail.

[shahed99]

Connecting Audiophile switches like SoTM or ER with the router - is everyone using SFP port lately as oppose to Ethernet cable?

[shahed99]

On 3/8/2020 at 10:32 AM, guiltyboxswapper said:

I did see this posted, infact, I brought the planet SFPs as soon as Emile mentioned them.

@guiltyboxswapper @matthias I've a favor to ask. Can you guys please share what fiber cable are you using between the planet SFP mentioned here? Do I need to use a attenuator given they are for long distance?

 

I've a Mikrotik router and SoTM switch and planning to use the SFP between them. Thank you!

[shahed99]

The other day I fired up my system after coming from work. And the sound wasn't the same; it was harsh and fatiguing. I was surprised why it was sounding bad all of a sudden. Then it hit me - the night before I swapped out my fiber connection between my Mikrotik router and SoTm switch with a Supra ethernet cable. But I didn't have time to test the configuration as it was late and went to bed. As soon as I put back my fiber connection - the sound returned; calmer, less harsh retaining the dynamics.

 

I bought the Planet Tech wide temp SFP last week. My single mode fiber cable didn't arrive it. But the single mode Planet SFP worked fine with my existing multimode fiber cable. The improvement was immediate. I noticed a wider soundstage, the sound was even calmer and relaxing. Once I received my single mode cable the next day, I think I noticed further improvement but I'm not sure.

 

I'm using a single mode fiber cable of length 5m. Should I go for a 10m cable as suggested by @Nenon?

[shahed99]

I'm a big believer in optical now since that above experience. In order to completely get rid of ethernet/copper connection, I need to find a suitable replacement of my JCAT NET card. Is there any audiophile grade PCIE optical lan card? I think startech makes one but not sure their quality.

[shahed99]

On 6/30/2020 at 10:41 AM, flkin said:

I purchased he latest version from Max Townshend and in some ways I received good results, but mostly it was again negative. What a downer - the anti-vibration and isolation theory was sound but practically it didn’t play out. Mids too forward and harder. It was supposed to provide immediate improvements - and without any breaking in time!

 

When I got to this part, my reaction was "no way!" lol. I'm also a recent owner of Townshend Podium. I bought them to placate my downstair neighbor mostly. It didn't help much in that regard. But the sound improvement was immediate and satisfying. These are the keepers.

 

Then I read the rest of your post. What do you know, I have used DIY Duelund tinner wire in the past as my previous speakers internal wiring! But have been using Western Electric tinned copper wire everywhere (IC + SC) for few years now. Duelund hook up took the concept from W.E. popularized by Jeff Day's blog. Anyway, those DIY WE wire replaced my $2.5k worth of interconnect and speaker cables and it was an improvement. Haven't tempted to try any other brands of analog cable since then.

[shahed99]

23 hours ago, flkin said:

Ah how interesting! I got there with the Townshend Podiums eventually after switching to tinned copper. It wasn’t an immediate step up for me but took some rearrangement of cables to get there. 
 

Right now I’m finding a naturalness and correctness to the sound that I haven’t heard before with the tinned copper loom. But one thing unusual I noticed is the daily variations of sound seems to be more than before. One day, it’s spot on sound and another day bass gets fatter or soundstage is smaller. I suppose it’s to do with breaking in as it’s happening less over time. 
 

You say you’ve switched over to $2.5k wires, is this still tinned copper?

 

I replaced the $2.5k worth of interconnect and speaker cables with western electric ones. The previous ones weren't tinned copper. Also, I think the duelund hook up wire takes a while to break-in.