750$ Intel NUC vs $6000 Aurender N200: I don't hear the difference

I took my Intel NUC, which has a custom analog power supply to RMAF one year. The Aurender was used at the show, but the system was set up using my NUC. Aurender lost big time, but it really had no chance. I use my NUC in an engineering and a recording system besides using it as a streamer for my personal stereo.

So, the real answer to what is better lies in the answer to the question "what software are you using". My answer, "the good stuff". This includes the driver for streaming ie USB driver software. I do not use the "free stuff". What's the difference, the clocking and the data stream accuracy.

I put these software packages on my engineering laptop, gamers would like it. It performs almost as good as my NUC as a streamer and will definite give commercial streamers a run for the money.

I have been using computers in my engineering, recording, and playback systems for years (Kleinbeck Engineering), 40 years to be exact. That was before IBM sold their first "PC's" and Sony made the first CD player. It makes analysis of the results much easier and much more reliable because I calibrate my systems to a known standards.

However, simplicity does have its rewards. Not a computer nerd, maybe commercial streamer is a better choice for you.

OP, do not be led to believe there is anything wrong with your hearing or impressions. You are not the first to report hearing zero difference between servers. I suspect many owners of high cost "audiophile" servers will denounce such a finding as nonsense.

I have long been skeptical of high-end servers. Some claim to use custom motherboards, but most use off-the-shelf processors from AMD and Intel. Their proprietary operating systems are touted, but most are based off Linux. I also see a lack of accommodation for upsampling, which requires a lot of processing power. Most servers are simply not powerful enough to do upsampling, at least beyond DSD128. Support for software like HQPlayer is also lacking.

If one is willing to do DIY approach, a SOTA server can be built for fractions of the cost. Take a Mac Mini or Intel NUC, boot it from Audiolinux (which has native support for Roon and HQPlayer), add a LPS, and connect it to your network with ethernet.

In terms of your problem logging in to Qobuz, your firewall may be blocking their site. I’ve had this problem and had to have Netgear tech support unblock it in my Nighthawk Wifi 6 router. My thoughts are that with a tube based DAC you probably won’t hear the difference between the lower noise floor of the N200 over your NUC, especially while using ROON because the tube noise is above the noise floor of the least capable source, namely the Intel NUC. I recently moved from a similar setup, got rid of ROON which has it’s own SQ issues and am now using a high quality streamer/server made by NOVATRON that has built in storage and runs on a linux custom processor and has exceptionally low noise characteristics, a highly accurate clock plus exceptional physical build quality with much attention to detail, yet only costs about $2K . This allows my (slightly) below premium grade Gustard X16 SS DAC (connected by USB)  to really strut it’s stuff. One would not have guessed a streamer/server could make that much difference but it really does. 

OP could be experiencing sin of omission deal, nuc vs. sin of commission, Aurender. Now is that sin of commission due to Aurrender or some other part of system?

One often doesn't perceive omission as a sin, and it may be preferable in some cases, perhaps in a lot of cases.

It comes from experience. I have never experienced anything in audio that could be 1) Verified , and 2) Not measured.  Usually #1 eliminates almost everything, and when it does not, #2 takes care of the rest. Every single time where I have heard, or experienced a verifiable sonic difference, there was also a measurable difference. Not just measurable, but above what are accepted limits to hearing sensitivity/discrimination, and significantly different.

I am not talking single point, THD either, but full scans, i.e. THD across frequency and output power, IMD with 2-3 points, and multi-point at multiple powers. Then again, just taking two time correlated streams (digital source only of course) can be highly revealing.

I think we both have respect for isolation, at least for the analog. I do believe a lot of the claims of "noise, RF, pick your poison" getting into the DAC clocking is for the most part suspect. If you know RF, you know that there is more RF coming off that trace 1/2" away then you are going to pick up externally. If the people who made those claims knew depth of D/A design, they would be far more worried about the D/A analog reference :-) ... but also not difficult to regulate for audio frequencies and eliminate noise.

I have convinced myself of things I heard that I later proved were not there. You are so hyper focused that something must be there that you convince yourself there is.

@cindyment

Thanks, and yes, either could impact sound, and neither is completely easy to 100% eliminate.  All the power supplies and isolation i refer to are custom designed and built by me - as are some of the interfaces and clocks.  So with luck, that's not the issue. I do believe i mentioned that isolation quality varies and I strongly prefer good transformers.

But what i experience is really not the issue. I wish i could say "do this, the rest makes no difference. But it does not appear to be that clear.  I am hoping to point to issues that others may experience, and why "cut and dry" technical answers give me pause.  I have spent my career helping giant tech companies overcome technical sclerosis that is bred by engineers becoming comfortable with one paradigm or another - things they are invested and expert in.

I suspect that we have a difference in perspective. You seem very comfortable with what you can measure.  I find Audio just full of the "unknown unknowns", things I can hear, but have difficulty measuring.  This is nothing new - once (40 years ago) clocking and jitter was totally swept under the rug.  We had the same problems in the 80s and 90s when we applied compression to both video and audio, often with studio processing as the consumer. They are not forgiving, nor are blue screens.  And traditional measurements either misled or indicated "it wont work" when in fact, it did.  My objective is to slowly learn what to measure, or at least what design concept correlate with good sound.

“if it measures good, and sounds bad, it is bad. If it sounds good, and measures bad, you’ve measured the wrong thing.” – Daniel Von Ricklinghausen, HH Scott to the Boston Audio Society, 1954

@Cindyment

Thanks, and yes, either could impact sound, and neither is completely easy to 100% eliminate.  All the power supplies and isolation i refer to are custom designed and built by me - as are some of the interfaces and clocks.  So with luck, that's not the issue. I do believe i mentioned that isolation quality varies and I strongly prefer good transformers.

But what i experience is really not the issue. I am hoping to point to issues that many may experience, and why "cut and dry" technical answers give me pause.  I have spent my career helping giant tech companies overcome technical sclerosis that is bred by engineers becoming comfortable with one paradigm or another - things they are invested and expert in.

I suspect that we have a difference in perspective. You seem very comfortable with what you can measure.  I find Audio just full of the "unknown unknowns", things I can hear, but have difficulty measuring.  This is nothing new - once (40 years ago) clocking and jitter was totally swept under the rug.  My objective is to slowly learn what to measure, or at least what design concept correlate with good sound.

“if it measures good, and sounds bad, it is bad. If it sounds good, and measures bad, you’ve measured the wrong thing.” – Daniel Von Ricklinghausen, HH Scott to the Boston Audio Society, 1954

It is not the issue. Burn in on a server , 150 hours.... come on. That is stretching the realm of believability beyond the usual in this hobby. Swamp land in Florida anyone?

I again spoke to the dealer. Asked him if he was sure if the unit was burned in. He got back to me yesterday saying it has only 60 hours on it.

Aurender says the N200 needs 150 hours to sound its best. Maybe that is the issue.

The unit is going back and I might demo it again in a few weeks after its 100% fully burnt in.

Do people realize that whether a computer use 10 watts or 500 watts, it still must pass the same radiated and emitted radiation standards? There is no guarantee the lower power one is less noisy either, though it is often easier to shield, but that does not mean that it is.

RF fields from unintentional emitters (computers, etc.) are fairly small, even on cables assuming you are using something that has met FCCICESIEC standards for consumer/residential. You don't have to move very far away from them until the fields are low, probably lower than a nearby radio or TV station. Those cheap meters pick up stuff, but you have to be quite close to pick it up.   And even if there, there is no guarantee it will get into the end equipment and most of you are just guessing that it does.

Every single process on motherboard produces noise. Every single cable, connector, LPS, switching power supplies, piece of equipment produces noise. Less than optimal home and equipment grounding, quality of AC are another source of noise. Noise is in form of EMI/RFI, also external sources such as wifi, cell phone towers, etc. I have Trifield Meter, can assure skeptics there is much RFI coming out of streaming equipment.

In speaking with guys at Uptone they mentioned speed of processors on motherboards, point being limiting time signal on motherboard important for sq. Incredible how far into the weeds one can go with this stuff.

I perceive server in much the same way as itsjustme. The majority of what you're paying for in off the shelf servers is r&d. Anyone with knowledge of computers and network can diy solution, the parts pretty similar. Implementation, the r&d of things is where differences are magnified. I recently saw, can't recall where, $35k server, amazing attention to every single little detail, quite possible this server would be revelation. Still, I couldn't account for $35k cost regardless of this maximum attention to detail.

Two considerations, performance and value, maximum performance doesn't care about price. Each of us has to determine value for ourselves. Ultimately, we all want maximum performance at minimum price. Determining maximum utility from streaming setup is so difficult, so many choices!

Wow, melton = melt down? Man take it easy. Don't have to be harsh on someone who is reporting their own findings. I think this is a great thread for folks interested in investing in streamers.

Can someone throw some light on why a dedicated streamer like Aurender/Lumin/Auralic might be better than a fanless NUC? What introduces the so called "noise" in the NUCs compared to the dedicated streamers. The reason I ask is because the streamers and NUCS both support some type of OS; which means the essential hardware has to be in place. And please note that this is NOT an "ASR type" question. I really mean to understand the differences.

@itsjustme 

With USB isolators, it is best to use a low leakage power supply, whether linear or switch mode. That could be why you are experiencing differences. Some of the USB isolators my also have built in DC-DC across the isolation barrier. The cheap ones do, but you can turn them off, but the circuitry may still be there. It is probably not an issue in almost all systems, but may be in some. No magic though, as you have shown, just common sense approach to the problem.

{. . . }  does not think bargains exist, they just think the expensive one is over priced. There is a difference.

precisely my thought.  Its true that 'too good to be true' generally isn't rue, but let's be real - what do you think is inside that streamer?  A computer.  Maybe a NUC.  Maybe a Pi.  Maybe something else. A LPS, a nicely isolated SPDIF and USB interface. All well and dandy - bu you are paying a lot for it and the fancy faceplate (generally the fancy package is the single most costly part of any high end component, followed by power supplies).

Now, i do think that you ought to be comparing the price/cost to a NUC, +LPS, +isolation between the NUC and the USB I/F and if possible a separate power supply for the USB.  +Excellent clock if SPDIF is used and in source master mode.  That begins to add up.  But not to $6k

The isolation may well exist at the DAC's input. Not all methods of isolation are created equal - i believe in using a transformer to fully isolate including split grounds.  I have also found, and cannot explain, that even with most "isolated" USB inputs isolating the driving USB from the computer/Pi still makes a small, but real, difference.  My only explanation is that the isolation is good, but imperfect.

I re-iterate that a 3- layer solution with server, bridge and DAC is my preference (Roon core, roon bridge, roon endpoint). It forces more isolation and moves heavy processing away from the audio signal (processing = power draw = power supply and ground noise).

The new Pi4s are leagues ahead of Pi3s in terms of USB bus powering BTW.  Alas they still draw a LOT of power at startup, as do NUCs.  Like 2+ and 4+ amps respectively.  So actually i power the Pi in split mode - noisy side with SMPS and audio side with LPS (~< 0.5A)

I myself have been hovering around the use of a Streamer Set Up as a Source to work with a Bespoke Built Valve Output DAC I am using.

To see a comparison that is showing that a Streaming Method has the Possibility to be closely compared to a much more expensive device when demonstrated to each other side by side is appealing to myself.

It is certainly enough to encourage myself to consider the method as a starting point and not have to part with the Monies required to acquire a much more expensive Branded Streaming Device. 

In the past I had looked into a NUC as I had seen it referred to as a method and I am sure there was a leaning to the NUC, to be used to supply the music files only and this could be done using the NUC RAM, which was claimed to be much quieter as a method. It is all in the past now and I may be a little bit off course on the method being referred to. 

I am yet to be convinced from my previous experiences of being given demonstrations of Streaming Devices that a Streaming Source is able to deliver in a manner that is on par with a well thought out Vinyl and CD Source.  

How the comments made by other contributors to this thread reflect on how the assessment of the comparisons has been evaluated are beyond my learning stage at present. I am happy to learn more on this if further description is offered. 

Reports like the one from @essrand about their experiences encountered are welcome by myself, and this thread has been a valuable discovery.

Do you really believe such bargains exist? I don't think so and neither does anyone else.

Everyone who understands the technology does not think bargains exist, they just think the expensive one is over priced. There is a difference. If I have a $1000 - 1 gallon pail of white paint and a $40 - 1 gallon can of white paint, and no one can tell the difference when they are on the wall, is the $40 can a bargain or the $1000 one over priced. Some just cannot except that one of the tested products is a $1000 can of white paint.

Burn in on a streamer?  What exactly will change w.r.t. what comes out of a streamer after burn in compared to before?

Streaming is not hard. Any modern commercial computing device does it well. Dac is where the big sound differences come in.  It’s been this way for years.

Take for example your common Roku streamer that cost less than $100. Does high res video and sound which is way more challenging than streaming say cd res audio alone. Case closed. You will find good sounding audio streamers at most any price point. Just because historically good quality music sources were expensive vendors leverage that to bring modern products to market. But the value proposition of high price streamers alone is highly questionable IMHO.

We should better understand this as one was better only in terms of subjective sonic preferences and system synergy. The system and owner ears/preferences wanted more full bodied warmth. His comments clearly said this. In audio, nothing is ever universally better. It often comes down to system and room synergy coupled with subjective sonic preferences. I owned the same Roon server and my Innuos Zen was a clear step upwards sonically. Just so much better to my ears in my system. No right or wrong here. Always best to hear a piece in one’s own system to determine for yourself. Glad the OP did this.

In a system with plenty of warmth and body and an owner looking for more speed, detail and resolution, the other server may have been preferred. Welcome to the wonderful world of subjective tastes and all manner of variables inevitably leading to differing conclusions!

The conclusions found in this one instance have absolutely no bearing on what you would determine with the same two units in your system.

@essrand I honestly don't care less what you think about me but you are not giving an honest in depth review. It's just you and your mate who came over on Saturday afternoon and chilled and played some tracks and thought "Hey man this is the same as six grand."

Do you really believe such bargains exist? I don't think so and neither does anyone else.

Anyway carry on in ignorance.

@sns That was the first thought me and my friend had. We immediately called up the dealer who said the N200 was in the showroom since July and has had at least 100 hours play time.

@jallan, please check my second last post, they were both tidal in the latest test.

@lordmelton It's okay if you don't approve my testing methodology of playing my favourite songs and some audiophile test tracks that I and a friend know in and out and A/B test the two servers. I am not sure how your method is different, plus I would test streaming because I want to listen to streamed music who cares if aurender sounds better on a USB stick if I only listen to Qobuz or Tidal.

Not sure why you are so angsty, did you just drop 10K on an Aurender or something? :)

Just reporting my findings in case somebody is in the same boat as me. My same friend who auditioned with me, is happy with the listening tests we did, because he was seriously considering picking up N200, and now I have saved him $6000.

I will keep an open mind and audition the N20 whenever my dealer has it available (soon he says), or other servers, maybe a 432Evo or Grimm or Fidata. But for now Roon ROCK it is.

This is quite amazing finding considering you're not using LPS on NUC. With LPS and fanless base I could more understand result. Are you sure the Aurrender is burned in, the analytical sound you speak of is common with lack of burn in.

Or perhaps results not unexpected. I have Trifield Meter and find things like displays, hdd, wifi, transformers give off horrible levels of RFI, the reason I like all separates. I'm careful with placement and provide shielding  around the worst RFI offenders.

mmmmm Petrus....nwo with that ANY server will sound better

@essrand ,

This appears to be a "you can never be woke enough moment", complete with the required insults.

Extreme positions and echo chambers. No matter what you do eventually you do someone will eventually try to ostracize you for wrong speak.

The Petrus comment just tells me that someone is offended that you were allowed in the club without paying the membership fee.

You obviously don’t understand the difference either. That first time wine taster was still able to tell the difference between the Australian and the Petrus. I’m showing a link that exposes that even very experienced wine testers will prefer other wines over Petrus. It could have been a vintage thing it really doesn’t matter. There is no objective best in wine tasting. There is a most objectively accurate in audio but there is no subjective best either.

@essrand I suggested that you do a simple test and you decided not to do it. You possibly can't explain away what you've done as a review and your reputable reviewer shouldn't put his name to such an amateurish ensemble.

Your sound comparisons are meaningless you don't even mention what you were listening to. It's pathetic gibberish.

Better off sticking it all through Cindy's multimeter.

I really don't believe you know what your listening for. Just like someone at their first wine tasting who likes an Australian Shiraz above a Petrvs.

If your DAC is competent then the result is as expected just as expected as the pages of posts trying to explain away what is the simple conclusion.

I question the Nagra implementation and conclusions on the analog side wrt it being the purest implementation as evidenced by the distortion spec, but it is more than low enough not to be a concern. Digitally the architecture seems sound so why would there be a difference? If there was a difference that means the DAC is inherently flawed. Why do so many people want to jump to the conclusion that your DAC is inherently flawed I have no idea.

@yyzsantabarbara I used to have a couple of MicroRendu. They are electrically noisy. I got rid of them because it just didn't seem worth the effort and I ended up changing the architecture of my system anyway. I saw ASR did some tests later that they are sensitive to the power supply. I would have ended up getting rid of them anyways because the firmware did not support any form of time synchronization between multiple units which I needed. However I was disappointed because I was really hoping they would have worked out.

Sorry, but your test is not under identical conditions.

Two nights ago I did an extensive test between Quboz and Tidal MQA on a decent live recording in 44khz/16 bit (Phish at the Berkeley Greek Theater in 2010), interleaving the songs on my Aurender N100H playlist.

The Quboz versions had substantially better dynamics and punch than the Tidal MQA versions, which sounded smoothed over and dull. This has been consistent with other tests that I have made of same resolution recordings available on both Quboz and Tidal. My conclusion is that you can hear the lossiness of MQA, and you are doing just that in your own test.

Use your ears...

Want better or different sound? Spend the extra money on better speakers, room treatments, make sure you have plenty of power, or play with your DAC / Pre-amp, get a subwoofer etc.

Computer has one job. Deliver bit perfect data. If you use a little DSP with Roon (as I do), make sure your computer/server/whatever you decide to call it... has enough processing power so it doesn’t bog down. Maybe spring for a solid state drive.

Love Roon however, I went in with other audio nerds early on for the lifetime subscription on a group/gift offer Roon had going. Saved some cash there.

My server uses Linux, and fortunately the mfg (Salk) provides incredible support and made sure the latest update would work as it was a bit more of a change.

It also won’t shock me someday if like most computers that I will need something new to continue to run Roon. This isn’t because of owning a particular brand, but after years of buying nice Apple (or insert brand) computers, they all seem to "outdate". (that is an entire other subject).  I hope this isn't the case, but it won't shock me.

But, feel free to drop $7500 or more on a dedicated audio computer if you like, it’s not my money! (but I would spend it elsewhere in the system...)

@essrand 

Thank you for your work and your post. I'm currently using Audirvana and found the same when I did a trial of Roon. However, I have not investigated ROCK and this looks like the direction to go for me in 2022.  Having over 9TB stored on an SSD drive now makes it worth putting together.  Cheers.

On some other comments, i know ROON makes a big deal about plenty of CPU etc. But it comes down to two questions:  1) do you employ DSP? and 2) do you simultaneously stream and synch multiple stream? If no, it runs on about anything (e.g.: i ran it for a while on a 2009 macbook pro laptop with  Core 2 Duo chip and also running plenty of other apps - ancient!).

Now that i employ a little DSP and sometimes synch 3 zones, i need more power. I have a quad core i5 NUC, a bunch of RAM (32? 16? forget) and two SSDs - one for Roon and one for music. Truth is that 99% of my listening is Tidal.

I would not expect a difference between the two servers.

That said, i would expect a difference between USB and COAX, although, i guess the Aurender has a very good clock (if its master).  I would also guess you can get an improvement by having ROCK --> bridge (e.g.: RPI or other with very good LPS and isolated USB output (isolate both signal and ground, separate power).

I do this, and have been for a couple of years.

Its often useful to also isolate the SMPS brick of the NUC from the rest of your AC with a decent passive filter.

@essrand  Good on getting together with your friend to sort it out.

One caution: you have a conclusion...but it is from a 'single' data point.

Hold on to the openness you've demonstrated.

I'm sticking with my Grace Digital Link...$179...doubt there is THAT much difference between it and any of these hi-end streamers. Certainly not a difference large enough to justify spending literally thousands more!

as a server importer you are wrong on a number of facts

it does not take 6k to beat a nuc our  a higher end server will blow a nuc away

we are roon dealers and we have the  nucleus on display a nucleus  is a nuc in a nicer case with or wihout an lps

our 5,500,00 432Evo high end model sounds far better and the 7,500.00 Aeon will shock you with just how good a server can sound and the sonic difference these devices can make

if you love Roon and really want to experience how much better your digital can sound you should try one of our servers

we agree with you Roon is the best

our servers were designed from the ground up to run Roon and they are full cores as well as being an endpoint

Dave and Troy

Audio intellect

US importer 432Evo music servers

So, me and an audiophile friend got together to sort this out.

All the caveats were resolved. Both NUC and Aurender used USB, both were fed by Tidal. Now the surprise was that NUC beat Aurender by a nice margin.

The Aurender sounded cleaner and more audiophile but more analytical and lacking lower end bass (hence cleaner we felt).

Roon ROCK on NUC had better bass and was more natural sounding. 

Go figure.

Hmmmm...now if I add a fanless chasis and LPS to my NUC it might sound even cleaner and beat Aurender in every which way.

The N200 is going back.

Anyone talking about noisy network or noisy USB etc are missing the point. This is a shootout under identical conditions. Roon + NUC vs Aurender N200 running Aurender software.

This thread might sound like heresy to the "source is as important as DAC" crowd, but I am as surprised as you are, and frankly quite disappointed.

Conclusion: Roon has done a fantastic job with Roon ROCK that has obviated the reason for a music server unless you want to spend the big bucks (>6000$ I am assuming, and even that needs some testing).

Also Roon has a UI next to none, a nice bonus.

Roon on mac mini was always meh, Audirvana sounded better, but not the same with roon rock.

Also: Am now thinking of buying roon permanent license.

In the same light I had an Innous player come in for an audition. It was played through USB. My normal player is an Intel nuc i3 with 8gb ram, using Audirvana 3.5 and out a Matrix USB card all attached to a LPS. 

The Innous was bested by a bit but both were fairly close.

Later I brought in a Bryston BDP-3 player and went through the same audition. This time the Bryston player won out. But again the difference was minimal. As a side note, the software although seemingly old school, is extremely quick, and bug free.

I would now like to bring in an Aurender  to try again, but I think due to their software issues I read about online, I really don't want to deal with those problems.

p.s. Introducing Roon never helps the sound quality, and it is why I will never buy it even though the interface seems to be second to none.

@djones51 

Most streamers, not all, don't have the necessary CPU and memory to run roon core. That's the main difference. I keep my NUC in my office and a streamer in the audio rack, both on the same network.  I have used a NUC directly connected to a DAC via USB. As long as you have a good modern DAC using asynchronous USB,  for instance a Benchmark DAC 3, noise shouldn't be a problem. 

Streaming into my Benchmark DAC3B in an all-Benchmark stack the differences between an Ethernet based streamer and a Fibre Optical based streamer was massive via DAC USB. My network is likely noisy with a lot of computers (about 10 machines for my work), network switches, etc...

I kept everything the same on the system except switching from a Sonore microRendu to an Sonore OpticalRendu. I also got a new network switch that had 2 SFP slots for the Fibre Optical. I do think the DAC3B was suffering from noise in the network. Or, the electronics on the microRendu was not as clean as the electronics on the OpticalRendu. 

been streaming (ripped music and/or Tidal) for a dozen years. When I moved from a Audirvana/Mac mini setup to a dedicated streamer/server, the Auralic Aries sounded better than the aurrender units and the DS Lightning software was also much better. At that point using USB to external dac. After a couple of years with the Auralic, got a very good dac with ethernet and went with Roon on a dedicated server using ethernet to dac. USB is a flawed interface and your better days use ethernet or i2s.

@essrand What I would suggest is eliminate the streaming first and judge each on their sonic merits alone. So record your favorite track, one that you have heard 100's of time onto a USB stick. It probably would help if it has acoustic guitar and voices. One of my references is "And You and I" by Yes.

If after comparing 20 or so times you can't hear a difference then there is no difference to be heard.

If you can then your router probably needs isolation or something else but do this first.

Good Luck.

@ghdprentice 

Agreed. Something is off, I can't figure out what. Am having another experienced audiophile friend come over this afternoon and perhaps we can sort it out.

My USB cable is Fidata: which is relatively unknown but a fabulous cable that beat Siltech and Audiomica cables in my system.

The Coaxial cable is Acoustic Revive.

TBH as an audiophile I am rooting for Aurender, not for saving bucks. I am looking forward to getting a server that will give me a massive upgrade in my digital rig. What audiophile wouldn't? Otherwise I have no place to go for a digital upgrade except some rabbit hole of ethernet cables and random network upgrades and ethernet regen and clocks and misc doodads. 

Your experience is very curious to me and completely different from my experiences with Aurender servers, I still own a N100 and a W20SE. They have completely out classed all PC based servers and other similar stand alone streamers. I would work with your dealer on this. This plain should not be true.

What is your incoming power like? USB cable. There is something very wrong. While I probably would have gone with an N10… yours should best a NUC without breaking a sweat. PCs are very noisy environments… regardless of what else is running. I recommend some real problem solving to find root cause… something is not right, although it does not stand out to me.

@lordmelton You got me confused with someone else, I have never posted about Aurender. Also I have no affiliation with audiotroy, he was just the first to respond to my thread with his promotion.

@auxinput My system is revealing enough, it can detect minor USB cable changes, leave alone whole component changes. Please google. Nagra Tube DAC is one of the top 2-3 DACs in the world right now.

@grannyring Agreed, which is why I put the caveats, but even though the cables are different, so is the output, it doesn't explain why there isn't a big difference in the sound. From what I read and all big noise about source being as important or more important than the DAC, I expected the massive kind of upgrade that you get when you go from a $500 Allo DAC (or a similar $500 DAC) to a Bricasti M3 (A 6000$ DAC). In which case relatively MINOR CHANGES like Tidal vs Quoboz, or USB vs SPDIF, as well as USB cables or isolation won't matter. But is a shock that even a minor change like a IC, speaker change would have more effect than changing my digital source.

BTW No Isolation on both servers, but Aurender has some decent footers, the NUC none.

I am still trying hard to get Qobuz working on Aurender, will report back or might get desperate and subscribe to Tidal to make the comparisons even more legit.

   In- have not been good at all ,I have spoken with several people have had horrible customer service ,with Innous,and  getting back to customers , that is one reason I have not considered them what’s up with them ,covid is no longer a valid excuse things are getting back to normal.

If you don't hear a difference then there's no difference to you. I wouldn't care about others opinions if I were you. I've tried streamers from about $4500 and down and heard no difference. Buy a streamer based on features (do you need AES, USB, etc.?) and usability. Someone is going to tell you it's your USB cable or your system (not revealing enough has already been covered above). Ignore them and enjoy the money you save. 

Most streamers, not all, don't have the necessary CPU and memory to run roon core. That's the main difference. I keep my NUC in my office and a streamer in the audio rack, both on the same network.  I have used a NUC directly connected to a DAC via USB. As long as you have a good modern DAC using asynchronous USB,  for instance a Benchmark DAC 3, noise shouldn't be a problem. 

Since you have Tidal access, can you switch to Tidal in Roon on your NUC, to at least keep the source the same?    I know audiotroy and others say USB is generally always best, but I hear decent differences between USB and SPDIF in my system so comparing using the same interface and digital cable with your DAC would at least remove one primary variable.

That being said, and I'm a believer in tweaks making a relatively and worthwhile difference in my system, but at an almost 10x price difference in that price range if I didn't hear a BIG difference between the two, I know what my response would be; it wouldn't be trying to tweak the 10x product to sound a little better than the NUC.

@milpai, I also run a NUC with Roon and it sits with all of my noisy computer gear, but I end up running the audio stream over a MoCA 2.0 coax network and finally over fiber into the listening room right up to 1M to my endpoint and DAC (Bricasti then Mojo Audio).    Maybe a $6000 server sitting in my rack would sound better, but maybe it wouldn't and I haven't felt like trying the comparison the OP just did :)  If my NUC dies every few years, $500 to $700 grabs a new one with up to date hardware and my music is still carried by light before hitting my DAC...