qute 2 vs NDX as digital transports

A)  It's unlikely that they all use exactly the same streaming module, the Ethernet interface ICs may well be the same, but the digital processing isn't.
B)  The power supplies aren't the same
C)  The digital clock arangements aren't the same
D)  The PC board layouts aren't the same

All this adds up to different levels of jitter and interference on the S/PDif output.

I've heard lots of people say digital is digital and the transport doesn't matter.  All I can say is that I have tried lots into my Hugo and they have all sounded different.  I did use my Qute 2 into the Hugo as a transport for a while and it was pretty good, however, my current Mac Mini/Audirvana was hugely better.  Can't comment on the NDX but there are plenty on here who I respect who rate it very highely as a transport into a dac

dayjay posted:

I've heard lots of people say digital is digital and the transport doesn't matter.  

Clearly they are somewhat ignorant of the details and/or otherwise ill informed - a nice bit of systems and crosstalk engineering theory should sort their education out...  

Simon-in-Suffolk posted:

dayjay posted:

I've heard lots of people say digital is digital and the transport doesn't matter.  

Clearly they are somewhat ignorant of the details and/or otherwise ill informed - a nice bit of systems and crosstalk engineering theory should sort their education out...

Sadly many of them wouldn't have a clue what you meant by that!

Exactly Huge, over simplification and marketing block diagrams taken as 'gospel' have got a lot to answer for ... and are the scourge for the armchair 'engineer'

Thanks, everybody, particularly Dayjay for sharing your experience!

... talking about block diagrams: In the NDS, the streamer module feeds via a buffer and digital audio isolation to the the SHARC DSP (with a local DSP clock), and it then goes to S/PDIF via another buffer (NDS white paper). A longer path than I thought! A pity that there's no white paper for the uniti qute 2 - I'd like to compare!

BTW: something must be wrong with the idea that a DAC such as the Hugo can compensate for some jitter in the signal, but what?

Best,

Stefan

Stefan

There are clear and audible differences between each NAIM streamer - both via analogue outputs and via digital ones.

 NDX is clearly the best digital transport if it's to be used with an external DAC (assuming that no 'sane' person would feed an nDAC from an NDS).

Adam

Stefan, in the Naim streamer architectures, the DSP micro-controller/FPGA  (SHARC) provides the SPDIF bit stream and/or the i2S over sampled bit stream for the DAC converter chips. (Dependent on config setup)

Either way the DSP clock, and it's regulation and performance becomes pivotal to the device performance as either a transport or DAC.

BTW just about every serious DAC on the market takes jitter out of the signal now, as they nearly all have their own clocks and long are gone the days of the DAC obtaining its clock purely from the source.. However cross talk is produced when one stage interacts with another. So jitter can be removed to be within a specific specification, but noise (or in system engineering speak an error) is introduced in the process of doing this. This error or noise can be proportional to the input pre processed signal.

Now better decoupling between stages reduces this error, but in practical systems its not reduced entirely... Hence you can usually 'hear' the effects of a poor transport in a reclocking or jitter busting DAC.

Simon

Many thanks, Simon, I guess by 'naim streamers' you mean ND5XS upwards and not the Qute 2? Actually, to my surprise there doesn't seem to be a SHARK processor in the Qute 2 - I couldn't see one! Well, this would presumably explain the sonic differences mentioned earlier...

Best,

Stefan

I have used NDX ND5XS & Qute 2 into a Hugo and/or 2Qute.

NDX any day of the week and twice on Sundays!

G

Stefan Vogt posted:

Many thanks, Simon, I guess by 'naim streamers' you mean ND5XS upwards and not the Qute 2? Actually, to my surprise there doesn't seem to be a SHARK processor in the Qute 2 - I couldn't see one! Well, this would presumably explain the sonic differences mentioned earlier...

Best,

Stefan

Correct, the UQ2, UL, U2 and 172 use a different processor (I believe they use a micro-controller).  But that's far from the only relevant difference.  I believe the power supplies and internal precautions against RFI (such as suppression components and the internal layout) are just as important as the digital processor used to extract the S/Pdif signal.

OK, thanks everybody, I'm convinced!

S.

P.S. In case you wonder where the shiny SHARC processor that is so easily spotted in the NDX series streamers (and provides DSD support, future room & speaker correction and all that) might be found in the 272: Fear not, it hides under the streaming board, and has juvenated and consequently lost its silverback outfit.

S.