I’m looking for thoughts/advice/answer for something I notice when streaming the library of my Uniti Core through my Lumin U2 mini. After reading many of the threads here I thought I’d ask directly.
My set up includes the following:
Naim Uniti Core outputting SPDIF via coax to a Lampizator TRP3 dac
Lumin U2 mini outputting SPDIF via AES to the Lampizator TRP3 dac.
I stream Qobuz through the Lumin and I normally listen to the Uniti Core directly played through the dac. I prefer the sound of the Uniti Core above the Lumin but for convenience I accessed the Core’s library from the Lumin to give it a try. Needless to say the sound quality was not as good.
Both the Lumin and the Core are connected to the same router, I assume this is how the data gets from the core to the Lumin for playback. Would the degradation in sound quality be a function of the router or simply the fact that the Lumin does not produce the sound quality of the Core?
I ask this question because the sound quality (core streamed through Lumin) is not even as good as when streaming the same title (same sampling rate etc) from Qobuz through the Lumin.
Not being a techie I can’t respond with regard to the router side of things, but what I can say, and there’s been a separate thread on this, on my NDX2 and Muso Qb the same track in FLAC sounds considerably more refined and organic from my Core over a Qobuz stream.
I’m sure some of the other guys will be better able to advise.
My simple thinking is when i stream Qobuz it may have come through various servers, cables etc for perhaps tens or hundreds of miles, picking up noise etc along the way. As opposed to my Melco server sitting one metre away from my streamer? Perhaps @Simon-in-Suffolk can give a technical view on it?
Gazza.. if it was an analogue signal, then yes, but as it’s a digital signal, it could go to Mars and back and it would be the same.
The data is encrypted and encapsulated like just for example your banking data when doing a balance transfer… if there was any corruption you would have random amounts in your bank account… the same principle applies to digital sound data sent my network … finally if there was even a single bit corruption it would fail decryption and would be rejected and discarded as effectively the network protocols would identify it had been tampered with.
So you can rest easy.. what goes in, comes out, or doesn’t come out at all.
Two aspects.. streaming service providers follow standards for LUFS processing… so the average loudness level of you internet streamed track, might be different from a download or cd ripped track… ie what goes in my be different in the above example.
Secondly, your streamer undertakes more work with internet initiated streams as they need to be decrypted compared to local home UPnP which uses very old tech and sent in the clear.
More work equals more processing which can mean more noise in a streamer.
That makes sense as the discussion at the last forum visit about the firmware that sounded off to some was due to the processor having a greater workload.
I would try a good Ethernet switch for beginning. EE 8 switch for example.
When i begun streaming, my Unitserve ( Uniticore now) connected to Nds in SPDIF sounded better than Nds streaming in UPNP from the Unitserve.
Adding a good switch and quality Ethernet cables reversed the things.
Thank you Lindsay- if I read this correctly your experience is that ripped files played back via the core sound better than files streamed over Qobuz.
Have you compared a streaming a title from your core (not played directly) to the same title streamed from Qobuz?
What I’m experiencing is that a title streamed through the Lumin via Qobuz sounds better than the same title when streamed from my Core. Like you- I’m not a techie but I would have not expected such a loss in sound quality when streamed directly from the core.
Your response to what @Gazza asked grabbed my attention here as I believe what I’m experiencing is the opposite of your example above.
Files played back from the Core are the best SQ overall. When streaming Qobuz through the Lumin the sound is good- slightly brighter than the Naim but nothing to complain about. When I loaded the Core library into the Lumin and played a title it went out the window- basically not worth listening to.
Both the Core and Lumin are connected via CAT5 to an Amplifi Alien router (mesh), that is the entirety of the “network”. Could this result all come down to the quality of LAN connections in the Alien?
@frenchrooster mentioned an EE8 switch- does one have to go that far to clean up the sound?
Thanks for the suggestion- I will read more about this switch. At first glance it seems pricey for the purpose it serves- does one need to go to this level to achieve the result? I’m assuming the same device alleviated the issues you described above?
The description above is factual. How that manifests in resultant SQ is entirely device dependent and setup and your personal tastes. Device drivers and software also pay a part on resultant SQ… and remember digital noise manifests itself differently to analogue noise, so you might prefer the effects of low level digital noise. (For example some types of digital noise are deliberately introduced sometimes to improve subjective SQ in some setups… ie dither)
So my description is not opposite to what you have .. it factually can’t be, but the resultant sound quality/preference will be product, environment and taste dependent.
But it would appear something is a miss in your setup as you appear to be describing ‘cleaning up’ the sound using a so called ‘audiophile’ switch. These shouldn’t really be necessary, they are very basic and simple devices simply undertaking basic electrical hygiene.
Good luck on hunting down what is causing your issues, yes it sounds like some UPnP or similar interworking issue in terms of SQ with your Lumin product and Naim in your setup… however almost certainly this is nothing to do with your local network per se.
To help put into context with the firmware I am currently using on my NDX2 the best performance in terms of sweetness of sound and imaging and layering (which is always a good indicator of optimum setup in my opinion) is best with Tidal Connect, followed by Tidal and UPnP and then finally Roon… this will be primarily driven by software in the Naim streamer… as previously that ordering has been different with different firmware.
Yes I’ve directly compared several tracks but most Stormy Monday by Cream from the 2005 Live RAH set. To my ears the local Core RIP is more organic. Interestingly the same applies on my phone with a FLAC file from my Core on FLACBOX very much better than Qobuz, through a Dragonfly DAC and the same applies with R3.
In fact after 15 minutes or so I have to turn Qobuz off and have closed my account.
One point on your system are you having to play the Core RIPs at higher volumes?
There are cheaper ones and much more expensive too. We are hundreds here to use quality switches. The sound is noticeably better vs direct to a router.
You can try at least a netgear switch with Ifi power ( around 100 euros). Then you have Silent Angel Bonn , 400 euros.
The best for money, the Cisco 2960 switch ( eBay, around 40 euros).
To your question- generally no, but that seems to depend upon the production of the CD that was ripped. Some older titles will require an increase in volume to achieve the same listening level, more recent releases are not an issue.
I don’t think that the Core playback levels are exactly 1:1 with that of the Lumin but whatever differences there are I’ve always attributed to the differences in output levels between the two devices (coax for Core/AES for Lumin).
I’ll make an inquiry with Lumin on this. Honestly it isn’t the end of the world because the simple answer is just use the Naim for ripped disc playback and the Lumin for streaming Qobuz, which is the specific task these purchased were intended for.
I do find it interesting though that some owners definitely notice these differences and pursue methods of improvement while others don’t seem to notice it at all!
Well the differences I suspect often down to differences in people’s setup.
I am not denying the effects of so called ‘audiophile’ switches on setups, after all I believe for hifi I was one of the first to discover the effects of switches on early mainstream consumer hifi streamers about 15 years ago and write about it on the web… before so called audiophile switches existed.
The benefits I had measured from subjective listening assessment were gained by minimised common mode noise, earthed powersupplies, and stable serialisation clocks, as well as in the network itself effective mcast filtering. Audiophile switches later started to reach the market addressing some of these aspects.. (although most if not all of the consumer products exclude mcast filtering as that requires more capable digital network circuitry, that you would otherwise get in a commercial grade product.)
Now most of these aspects can be achieved with a higher spec commercial grade switch and improved further by seperating the streamer from the DAC… but in the absence of this an audiophile switch can help reduce noise from twisted pair Ethernet into the streamer. As I have said before for very low noise audio, twisted pair Ethernet is a challenging medium and not really optimum.
So if i understand correctly, for future application, I should be sure to have a good switch and from that switch a non twisted pair Ethernet cable to my streaming source?
A good switch, which I suggests is mains powered and earthed is a great start.
All electrical Ethernet is twisted pair.. it’s how it works. Alternatives are fibre Ethernet , which Naim doesn’t currently support and wifi. The latter can have many advantages but is dependent on the streamer implementation on SQ. in my experience some streamers have had lack lustre wifi implementations.
This is one reason why it can be beneficial at looking at a streamer and a separate DAC connected via USB Audio or SPDIF.