Unitiserve vs Mac Mini

I have a Mac Mini M2Tech Hiface and Naim DC-1 and have compared it`s SQ to a US using BNC dig out with similar cable.

US is a much better source and it`s User interface is much better than Mini (both using iPad)

I cannot comment on NDX, they are meant for different use.

Best

If you are using the asynchronous USB input on the DAC I would expect the Mac Mini performance to be better than that of the US.  The jitter performance of the US digital output is not stellar.

Hi Jan

I use a Wireworld Supernova Optical IC between a Mac Mini and the Naim DAC/555PS. This sounds fantastic. The Mac Mini has 256 GB SSD for OS X Lion with iTunes and Bit Perfect. It has a 750GB hard disk for my Music. 

I cannot ell you it sounds better than a US, but it loses nothing to digital output from a UQ. Hence based on what you told me before I am inclined to guess it sounds better than US S/PDIF out. (i.e. if MM = UQ and UQ > US then MM > US). 

I have never seen any jitter figures for the US or UQ. The Mac Mini is very low. Using a high quality interconnect is essential in my view and the Wireworld Supernova Optical IC is the best one I know of. I couldn't use my TCI because they don't do with a suitable connector to I had to buy a new cable. This reduces/removes RFI/EMI so it is not just about low jitter. 

Using a USB/S-PDIF convertor is unnecessary. The MF V-Link works every bit as well as the hiFace and INT202 (both of which I tried in a previous configuration). The MF V-Link makes no difference to the SQ with a new Mac Mini. I suspect the new MF V-Link will sound slightly worse because it uses coaxial without galvanic isolation. Although the hiFace is good, I shy away from it because it loads another driver on the Mac. Whatever Apple has done to improve the Mac Mini, I'm impressed - new S/PDIF out chip and no sharing with CD so I read. It is quiet and it does exactly what I want (with exception of play CDs directly). 

I have no evidence to suggest the US would not sound great and if Naim could put it forward as genuine CD transport, it would interest me. It has a good quality TEAC drive in a low vibration chassis and so Naim has no excuse for not giving it good digital outs. Perhaps they do and the rumour mill has taken a grip. Of course, when Naim tells me it is not intended to work as CD transport then it hits home. So if Naim would like to say - have a listen you have got it wrong - then I'll take heed. 

However using my logic above and what you've told does seem to concur with this view. 

All the best, Guy. 

Thanks Guy for the two-in-one reply. Regarding the reputation of the UServe dig out, my hunch is that it's the rumour mill, but I have been wrong before on one occasion and many more times after that. So, armed with hope of getting to the bottom of this, I've just written to Naim support for clarification.

On a side note, I did get to test the shock resistance of the UnitiServe ; one of my Mordaunt and Short Pageant 2s (with rebuilt crossover and no ferrite-core inductors in sight) was knocked off its perch onto the Serve by a mattress (don't ask). The Serve stuttered a bit, regained its composure and carried on as though nothing had happened. Not a glitch since then, just a dent in the M&S. Gotta love Naim.

Jan

AFAIK the UnitiServe is based on a PC-compatible motherboard running embedded Windows XP with PC-level components (including commodity S/PDIF outputs).  I wouldn't expect this configuration to deliver low jitter.

Input signal jitter is irrelevant with the Naim DAC as all inputs are asynchronous.  The situation is different with the M2Tech Young DAC as only the USB input is asynchronous (it uses HiFace technology).

> Input signal jitter is irrelevant with the Naim DAC as all inputs are asynchronous.

Why does Naim say the CDX2 is a better transport than CD5X, which it definitely seemed to be when I heard the two at a demo. Why does the new Mac Mini sound better than my old Apple computer. Why ..... 

I know what you are saying about the Naim buffering technology, but ... as the Who once sung I Can't Explain. 

I've never heard a Young DAC, still never mind, I'm happy enough with my Old Naim DAC 

Thanks for the US info - interesting 

So the S/PDIF out just happened to be on the US motherboard rather than being a high quality Naim digital output, this would explain why Jan thought the UQ sounded better (I think). Shame Naim didn't do something about this, as it could have been a great one box transport. Still let's see what Naim say to Jan. 

All the best, Guy

I think the UnitiServe was focussed at being a server rather than a high-quality digital source.  It takes quite a lot of effort to implement a stable, low-jitter clock such as found in the nDAC.

BTW - out of interest, what does the SSD do in your Mac Mini (apart from making it boot quicker)?

If you're using the Naim DAC then it's asynchronous and ignores the jitter in the US output signal - so I'd expect your SQ to be excellent. 

So are you saying all transports sound the same through the Naim DAC then .... ?

It is not my experience ..... or that of some others

Maze would you say US into Naim DAC is as good as or better than CDX2? Thanks, Guy. 

I'm not saying that all transports sound the same, but nobody has ever given me a convincing explanation of why they sound different.

There are two paramaters in digital audio - timing and data.  The NaimDAC takes care of the timing and getting the right data is trivial.

"Naim HQ" don't usually answer such questions about Naim hierarchy on this forum. However, Phil Harris had explained once in one of the threads, why it is technically likely to find differences btw dig-outs of various Naim devices.

Personally, I had found US dig-out not exactly up to the level those of HDX or NDX, lacking just the last bit of microdynamics. But, this is just relative. I can only confirm that US is a marvellous transport, sounding much better than many CD players I've experienced. I hope to compare it to new Macmini as Guy suggests. Maybe I can loan a Macmini from a friend or something.

A couple of weeks ago I demoed the HDX player server :

(1) Play the CD directly

(2) Rip the CD and play the ripped data from the server. 

My impression was that method 2 gave better results. Not night and day, but better clarity/separation. And that was the impression of all 3 people in the room (one of them an audio dealer).

So it seems to me that source still matters. Or at least the method by which the digital data is delivered to the DAC matters. 

Why the difference?

Let's look at a simplistic diagram that may help understand the parts. 

Let's assume that the incoming CD bytes (solid blue line) are the same as the files bytes (solid green line). i.e. that Ripping has produced correct sequences of data. Don't really care of the container format here (AIFF, WAV, FLAC, etc)

So :

- CD data for track 1 (solid blue line) will be 1,2,5,8,12,3….

- Decoded data for HD track 1 (solid green line) will also be 1,2,5,8,12,3….

Then :

- The transport delivers data to the DAC (dotted blue line). 

- The Server (Computer, NAS, Naim Server) delivers the data to the DAC in one of 3 ways. USB, Optical, or Ethernet.

It is possible that some sort of noise or distortion is accumulated to the digital signal during the transfer of data (dotted lines). In other words the data delivered to the DAC from the 4 possible methods is not identical. 

Another factor that may affect SQ is the DAC interfaces.

Could that be correct? Could that be measured?

Interesting thread but a couple of points spring to mind:

mind:

if you are considering userve vs new MacMini you need to bear in mind it doesn't have any kind of optical drive so you need an external drive or another pc to rip cds 

@YanC - I also find on my Uniti that rips sound clearly better than playing the CD - my theory is that playing a CD "live" involves lots of error correction etc. and thus the output from the transport is unlikely to be bit perfect; plus having a cd drive spinning away in the same box may well be causing all sorts of EMI, vibration etc.  

I don't believe the data's being corrupted - it is only running at kHz and computers successfully handle data at GHz.  As long as you can discern a 1 from a 0 then noise and distortion on digital data signals doesn't matter  I also don't see why the DAC interfaces would corrupt the data - again, it is a really low-speed data stream.

Maybe corrupted is the wrong word. What about polluted or colored.

A lot of processes running in both Transport and Server boxes. Mechanical and Electrical. Could it be that the signal in the dotted lines is colored by them, whilst it is delivered to the DAC. Hence the audible difference.

That would also explain the current trend to eliminate some of these stages by making all in one boxes, including amplification that comes after DAC.

Edit: 

Data going to the DAC's input is no longer binary (0/1), but 16 bit or more.

The beauty of digital data is there is no concept of polluted or coloured - like pregnancy, it either is or it isn't.

well, I am only trying to understand in layman's terms what could cause the audible differences. 

Digital data is stored as 0s and 1s on the disc, but, if I am not mistaken, this is not what the DAC sees. Data read from the store will be converted to 16 bit numbers. As far as the computer is concerned these numbers can be amplified (volume control) as well as altered (equalizers). 

But even if no amplification or equalization takes place, the signal itself that transmits those numbers may be polluted. So what the DAC sees maybe digital but not always the same from the various sources.

Just a thought.