Recent HiFiCritic review of ND5XS.....

I can certainly understand why someone might use a CD player, if they were moving to streaming and / or wanted to able to play CDs that friends bring round. As to why adding a CD player would sound better than streaming I have no idea.

Hi HH,

Well yeah, I too can see why someone may have both a streamer and CDP, especially if gradually transitioning from one to the other.  But an owner using the ND5's DAC implies that the CDP in question is a lesser quality machine than the ND5, which makes it even more surprising/unlikely that such a set-up would sound better than streaming mode.

I can understand why in the context of a test they might wish to hook up another digital source via SPDIF, to get a feeling for how the DAC part performs in isolation.  And if that source happened to be a high-end CDP then the outcome may not be a surprise.  But I can't see in the review what the CDP actually was.....

All in all, the test results/findings seem to be somewhat puzzling.

I put some WAVs from my Unitiserve onto a USB stick, and played them through the front input in my Superuniti. They sounded better than UPnP. Then I got my act together and sorted out my network. They no longer sound better.

;-)

I guess my expectation would be they'd sound roughly the same, neither one notably better or worse.  An interesting experiment to perform (with any streamer)......

Ha Mike,

He's a writer I'm actually unfamiliar with.  Is he renowned for being, shall we say, eccentric in his opinions ?  ;-)

As a general comment to the review (and presumably therefore to HiFiCritic mag overall) awarding points for sound quality seems altogether strange.  Its not an exam, is it ?  I can only guess at how they would achieve this -  perhaps by having a range of specific criteria, rating a score out of 10 for each, then aggregating the result.  But surely there would be scope to get a horribly wrong outcome with that, eg a lower scoring product sounding decisively better overall. 

Or perhaps he just plays the music to his dog, and then checks the howlometer......  ;-)

Hi karlosTT,   sorry I screwed your post order up,  but I have added something 

I am not sure eccentric is the right word, he is a journalist & needs to sell his work ............  

-----------------------------------------------------------

 .............  its Martin Colloms remember.  

Re using a .wav (or .flac) on USB stick .........

I've found this a good & easy test of how your setup is connected, wired, wireless etc. & used it on my system & played with a few things:  

My system is all wired ethernet & centred on a LAN switch,  I cannot hear any real difference,  but if forced to make a preference I would say USB is 2nd.

1st test was wireless - I removed the ethernet between NDX & switch,  all the rest was wired.  although is was OK 99% of the time I did get a 24/192 stutter breakdown, 24/96 & lesser files were all OK,   SQ was not as good as USB,  but I was surprised how good it was,  maybe a the few metres between the boxes helped.

2nd test was back to all wired,  but this time I discon'd the switch & hooked the ethernets thru the BT wireless hub. Apart from not suffering the 24/192 stuttering, I would say it is not as good as wireless.

On my Qute there is a noticeable difference between the USB and UPnp inputs.  I have configured this many ways, including both products (UPnP Server & Qute) next to each other with a Switch between Them.  I am about to test a crossover cable directly between them. 

In every case the USB sounded better.  Bass was more defined, and USB had a bit more attack.  

I can well appreciate the use of the USB input as a 'constant' in such network testing.  Which by the by serves as a sharp reminder I need to get a proper switch of decent quality, as I presently use the wifi router for this purpose, and know its a piece of cr@p.....

In every case the USB sounded better.  Bass was more defined, and USB had a bit more attack. 

Interesting Brubacca.  So Martin Colloms (the reviewer) was not completely alone in his findings.  I wonder if Naim have an official stance on this.  If they do I'd expect it to be something like "the USB input is there for convenience, and isn't intended as the primary source".  Though I may be wrong.

In one sense there is something delightfully simple about the idea of a thumb drive 'source'......

  Apart from Qute,  what make/model etc. is the UPnP, server & switch ?? 

Not sure about the Qute, but the Naim streamers buffer the file data, into internal memory i believe, first before playing. So surely there should be no difference between UPNP and USB. Or am i missing something.

Graeme

In a way. The streamers (and Naim DAC) use a RAM buffer as a way to try and eliminate jitter in the incoming data stream (data clocked in at one end tied to the input clock rate is then clocked out the other end by a very precise internal clock) but they don't load a whole track before playing.

James

But if it is loaded into memory and that is where it is played from, all be it a part at a time, then surely the way it gets into RAM should be irrelevant.

Although saying this i know network cables make a difference to my NDX, so i've kinda answered my own question there

Good questions Foxman, or maybe better say "musings"

I am with you on this,  I wonder what is a USB memory is & if it performs the same in replay situations as a buffer,  according to the Naim white paper block diagram,  both ethernet & USB go into the streamer module & that is upstream from the buffer.

Then what about those other cable anomalies, near & far end & alien crosstalk, skew delay & return loss.  Do these anomalies get carried into & affect the buffered data,   or is the buffer data truly a verified transmit/receive confirmed copy of the HD data ??

Whilst I am fascinated by this,  & I expect we will get lots of theories, some more expert than others, I expect they will still be theory; meanwhile I am just as happy that my system replay is the same on USB as it is on ethernet - at least according to my Mk-II audio receptors

The ethernet cable "difference" argument was recently theorised in Stereophile (or one of its affiliates) to be down to noise rejection. 

Presumably a thumb drive could not inject noise -  unless it acted as an antenna somehow ?

My experience with a ND5 in wireless mode is similar to Martin Colloms.  The overall sound quality of the streamer/dac is good but it can be improved with a power supply or my case, a dac.   The tradeoff is an added box, redundancy and a greater cost than the NDX.   Yet even with enhancements, the ND5 may not match a good CD transport.  Or so it seemed to me last fall when comparing the following sources:

(1) ND5XS

(2) ND5XS transport mode (V1 dac with DC1 cable)

(3) Sonos (W4S mod) transport mode

(4) Rega Apollo-R transport mode

When used with a Naim XS-2 (70 watt version) and Harbeth P3ESR speakers, the “naked” streamer fell short of configurations (2) and (3).   But what surprised most was how much more dynamic, detailed and natural the sound became with a CD player.   Unfortunately, the player had some disk reading issues, so it and the ND5 found a new home.   I continue to search for a slightly used CD5XS, but they are as rare as a nun in a bikini. 

Netgear GS105 Switch, also tried connecting both to TrendNet TEW-680MB Wireless Bridge, also tried connecting network through powerline ( no hardwire at connection).  Used the Powerlines to connect to network then switch for both qute/ uPnp server.  

Servers-

J River on homebuilt silent PC running media Server (latest)

Asset

logitech media server ( running on homebuilt pc and two different ReadyNas Duo)

built in uPnp Streamer on ReadyNas Duo v1 and new 102 Model

Plex on ReadyNas Duo

Daphile PC running on Homebuilt PC

Vortexbox

there was at least one other uPnP server i tried and can't remember

I also tried this (Flac, USB vs UPnP) on my SU.

Result, no difference detectable by me.

So basically it appears you are running a ReadyNas Duo thru a Netgear GS105,  is that correct ??

Why all the other stuff, is that not adding a confusion factor??

I have a ND5 XS, and, initially yes it sounded better when using a USB memory stick.

Then I sorted my network out...

Now it sounds the same via the network or the USB port (and in both cased better than my via my CD transport).

I suspect he hadn't paid enough attention to his network (after all bits are bits aren't they, so it can't make a difference except for when it does).

My trials have been an evolution over a long amout of time.  About 2 years, which led me to get the Qute in the first place.  I'm not quite sure how I could have made my network better.   I have spent considerable time/money fooling around with it. I am quite happy with my direct USB connection.  Other than my next trial of crossover cable directly to my Homebuilt music server, I'm not going to fool around past that.  I also do not use the ReadyNas anymore.  The Homebuilt server has more power.  

Running a dedicated ethernet cable to that location is not practicle at this juncture, considering i am very happy with the USB, which was not my intention to fix or debate.  Just state that I agree with the reviewer on this issue.