Comparisons with HDX, UnitiServe, nDAC, 555PS and NDX in an audiophile environment

Posted by: aysil on 16 January 2011

I’ve been an audiophile enthusiast for more than 20 years, and during the last years, I was lucky to be able to enjoy an ultra-neutral and ultra-transparent system. Lately, I was again attracted to equipment of Naim, a brand I always admired very much, because they seemed to suggest a way of introducing new formats to our music system without being dependent on a turned on computer. I decided to purchase HDX, UnitiServe, nDAC, 555PS and NDX (has not arrived yet) to do some experimentation. I am new to this forum, but I spent days and many hours reading the many discussions, and I decided I could share my impressions:

I started by ripping some of my CDs to the on-board hard-drives of HDX and UnitiServe and gave all devices around 300 hours burn-in time before any auditioning. After the burn-in period, I ripped some more CDs.

HDX+nDAC

My first impression with HDX connected directly to the preamp was that it had a very decent and enjoyable sound comparable to good CD players of the same price range. I was curious how much the nDAC would improve the picture. Comparing HDX alone with HDX+nDAC, it was obvious that both devices were from the same house with almost the same tonal balance – nDAC slightly on the warmer side. But the improvement with nDAC was significant; there was just more of everything: more silence around the music, more detail, more tone colour, more air… And the highs came much more smoothly from nDAC; although the treble is more recessed in HDX, it has a harsh quality which can sometimes be very annoying.

HDX+nDAC vs. HDX+PS

I wanted to find our which one is a more meaningful upgrade for HDX. After a few trials, it was obvious: although PS did improve the sound of HDX, with nDAC the music was simply more vivid. nDAC (or another high quality dac) is a priority upgrade for HDX.

nDAC vs. my reference DAC

Believe me, I tried quite a number of very exotic dacs (some of them were very bad, btw). I can just confirm nDAC is one of the best dacs ever made and it can compete well beyond its price range, even more so if paired with the external PS. Is it the end of the road? Definitely not, there are a number of better DACs, but many of them at extremely higher prices! The DAC I had chosen some years ago through careful auditioning remains imo the best dac on earth. I wondered if nDAC could approach that sound. I made direct comparison feeding them from the S/PDIF output of HDX: my reference dac positioned all instruments so realistically in space; you can almost feel the air between each individual instrument even in the same group and all instruments have the right size and body. This almost palpable quality paired with a wonderful fluidity makes the presentation more realistic and musical. The nDAC remains somewhat edgy in comparison.

HDX vs. UnitiServe

For most of us the internal dac of HDX is futile, and we are attracted to UnitiServe as a less expensive alternative ripping device, assuming their digital outputs are identical. Well, I have bad news: their digital outputs are far from being identical; they are in fact two different worlds! Through my reference dac, the S/PDIF output of UnitiServe sound nervous, rough, and the micro-dynamics more flat (in comparison to HDX). The HDX has a more relaxed presentation, more resolved, better flow, and more precise rhythm. I tried to find out if the difference lies on the hard disk processing or on the S/PDIF output stage of the two devices. Having them both on the same network enables them to detect each other and play FROM each other’s hard disks. The sound of the files on UnitiServe hard disk was better through HDX output than directly from UnitiServe. Moreover, UnitiServe output sound better playing files on HDX than playing own files! I think this shows that BOTH hard disk processing and digital output stage are superior in HDX.

Well, we know for more than a decade now that bit is simply not bit – even “perfect bit” is not – and the digital music signal is an EXTREMELY FRAGILE signal. So, such differences should not surprise us. There are a lot of precision and timing issues associated with the digital music signal and jitter is only one of the problems. Can we blame Naim for this sonic difference? Look at the immense price difference. It is apparently not only for the internal dac.

I was not able to compare the UPnP output of the two devices with their S/PDIF outputs – I am waiting for the NDX! I also don’t know yet if the quality difference between their UPnP outputs will be as pronounced. I am pretty sure, though, there will be some difference.

THE MOST EXCITING PART (for me)
HDX vs. CD-transport

When I started this experimentation with Naim devices, I was REALLY hoping, for the sake of convenience, that the ripping solutions would sound better than my CD transport. This was of course an unfair competition for the HDX. This 66 lbs. heavy CD-transport is engineered to extract as much information as precisely from cd as possible. But I’ve heard so much word about the “sonic advantages” of ripping that I was tempted to make a direct comparison. No, HDX cannot come near. The CD-transport is in better control of all the micro-dynamics of the music giving a more realistic representation of all the different effort qualities and conveys the real intention of the musicians. It has better definition and gorgeously natural tone colours.

Does this mean ripping is inherently worse than cd spinning? Most probably not, but it means that the ultimate hard disk player has not been made, and that computers in general still need perfecting to be able to serve as ideal music sources, with more attention to isolation, vibration control, internal wiring, metal alloys in disks etc. HDX should only be compared with CD players in its own price range. There is no sonic miracle in ripping; it has convenience advantages for things like multi-room etc. I will go on SPINNING. (cd and vinyl!)

MY CONCLUSION
One can argue that all my comparisons have been on cd format, and that higher resolution formats available for download sound better. I am aware of the potential there. I am also very excited especially about direct studio recordings or the direct transfer of the master tapes on high resolution formats. But, my comment is still valid. If an environment (the computer) is not getting the best out of the cd format, it must be far from doing justice to higher resolution formats. I appreciate all the people who give a tremendous amount of effort, through careful component selection and so forth, to transform their home computer to a valuable music source. They are the pioneers of the field. I personally prefer to wait until Naim (or another respected music equipment manufacturer) makes a dedicated music computer which bypasses the general purpose computer already from download phase, and which is engineered to audiophile criteria.
Posted on: 16 January 2011 by Mr Underhill
Dear Aysil,

Welcome to the Forum, and what a good contribution.

I have heard the 555 in comparison with two computer sources, and it was undoubtedly better.


I would love to know what your CD player and DAC are! But respect your choice of boundaries when using a commercial forum.

I was disappointed that you didn't use streaming via upnp. Will you do a comparison with the three - once the NDX arrives?

I was pleasantly surprised by the increased quality of what I was hearing when I switched from direct play to streaming.

Look forward to your future posts.

M
Posted on: 16 January 2011 by Tog
Hang on -

the 555 is a colossal amount of money and represents the final evoluton of the soon to be extinct CDASARRUS - so it should sound pretty amazing.

The UnitiServe was designed as a UPnP server - the fact it has digital out is a bonus -

Digital streaming / computer audio is in its infancy.

Lets compare like with like.

Tog
Posted on: 16 January 2011 by matpip
Very nice reading, thanks!
Out of curiosity, did you really purchased material for more than 15K GBP just for testing/experimentation?
Posted on: 16 January 2011 by aysil
Dear M and AllenB, as this is a forum for discussing Naim products, I thought it would not be appropriate to name and discuss about my other equipment. But I am ready to communicate on those if you send me e-mails to aysil31atyahoodotcom

Dear Tog, you are right, and I said it, I was being unfair by comparing equipment with far more expensive counterparts. But my emphasis was exactly what you are saying (and what I learned through these trials): "computer audio is in its infancy" It should not be presented as the ultimate audiophile solution.

I am also curious about streaming, and I am enthusiastically waiting for my NDX! I am just more cautious now knowing this infancy situation.

Comparing HDX with UnitiServe is a very valid one, though. The question was raised many times on this forum.

I don't tend to believe CDplayers/transports will get extinct, nor are turntables getting extinct.
Posted on: 16 January 2011 by Hook
Hi Aysil -

Welcome to the forum, and thanks for the interesting post!

So where are you now using your 555PS? Have you moved it to the DAC? Will it stay there, or move to the NDX when delivered?

At a recent show, a dealer paired his 555PS with the NDX because he said it lowered the amount of noise being sent upstream to the DAC. On the other hand, I cannot imagine ever separating my DAC from its 555PS!

Thanks for the clarification.

Hook
Posted on: 16 January 2011 by aysil
Hook,
555PS is connected very happily to nDAC at the moment. will start experimenting again when NDX is delivered.

I agree with you: PS would be a significant upgrade to any device. Even power-cords have a tremendous influence on sound. In fact, I always experienced power-cords (together with digital cables) have more influence than interconnects or speaker cables. When you consider that the power is not just there to "power" the device, but to provide the 'material' from which the device makes the music... Better raw-material, better end-product!
Posted on: 16 January 2011 by aysil
Yes I can do that sometimes. Different priorities; therefore I don't have an expensive watch, and my car is cheaper than any of my cables in my music system.

I am sure these wonderful Naim equipment will easily find a good place in my office, in a friends house, or find a new owner, even if I decide not to use them in my main system.
Posted on: 16 January 2011 by Dungassin
I don't think there is any problem in you stating which CD transport you are using. After all Richard Dane (one of our moderators) instigated a long-running thread on various transports with the nDAC. Winker
Posted on: 16 January 2011 by js
Thanks for the input. I'm actually more curious to the rest of the system used for comparison and if any extras like line conditioners video bits etc. were anywhere to be found. I suspect not but an overview would help perspective.

Depending on associated bits, it's sometimes helpful to remove another dig or vid item from the mains before a listen to another as some already contain parallel conditioning or create noise at their power input that can affect other items hooked to the mains. It can sway an evaluation but I suspect you have already been down this path. Winker Did you play with the ground switches? Has the DAC been used for a few weeks?I also hear differences in the various streamers and like a good transport but not as much as some of our HiDef sources. We use our old CDS-1s for spinning a dig out.
Posted on: 17 January 2011 by AMA
Hi,aysil! Wellcome on the forum and thanks for sharing your findings with Naim community.
quote:
Dear M and AllenB, as this is a forum for discussing Naim products, I thought it would not be appropriate to name and discuss about my other equipment.

This is wrong -- many of us make complex systems with occasional non-Naim inclusions. I use non-Naim transport and non-Naim speakers which makes my mostly-Naim system sound in the way I LIKE it.
Pls, feel free to publish your setup in your profile.
quote:
Through my reference dac, the S/PDIF output of UnitiServe sound nervous, rough, and the micro-dynamics more flat (in comparison to HDX). The HDX has a more relaxed presentation, more resolved, better flow, and more precise rhythm.

This finding is interesting and quite logical. But I guess that both HDX and UnityServe have low output jitter -- at least low enough to be re-clocked by nDAC with bit-perfection so when played through nDAC they sound absolutely identical.
Posted on: 19 January 2011 by Rosewind
Aysil.

A very nice write-up indeed! Please share info on your transport and dac.

Best wishes,
Peter
Posted on: 19 January 2011 by aysil
I made the HDX vs. UnitiServe comparison with my previous DAC first because this was personally more relevant for me, and second because I felt this DAC would be more revealing. It would be interesting to do a revaluation this time with the nDAC to see if it would "hide" some of the differences. At the moment I am travelling but I will do it as soon as I am back home. I will also call my friend with whom I did the first auditioning together. We heard the same differences and they were quite obvious. Now, I can do only some guesswork, if you allow me: I guess there will again be some audible difference in my system. nDAC introduced an original method of REDUCING jitter and noise. I am not sure if this means that it was designed for making up for the shortcomings of inferior source components. Let's see, we can discuss the implications.

Reference:
the statements from Naim that the dig out circuits and PCI board of the Serve and the newer / upgraded HDX are ostensibly the same.
I don't know what the differences in the inner topology of the two devices are. CD-drives used for ripping seems to be different, the isolation level between those supposedly same components may be different and many others... But there are two very obvious and significant differences between UnitiServe and HDX: The power supply and the mechanical properties of the enclosure case. We know that both have tremendous effect on sound, also on the digital domain.
Posted on: 19 January 2011 by aysil
Btw, when I said mechanical properties, I remembered that I tried Marigo Mystery Feet with each Naim equipment and the improvement they provide is just incredible. I can guarantee that they  are worth the pretty high price - an upgrade to a higher range. The other feet I like are the Shunmook.
Posted on: 19 January 2011 by Hook
Reference:
But there are two very obvious and significant differences between UnitiServe and HDX: The power supply and the mechanical properties of the enclosure case. We know that both have tremendous effect on sound, also on the digital domain.

Hi Aysil -

There have been a number of threads, some of which went on for months, that discussed the topic of whether the buffering/re-clocking architecture of the Naim DAC results in sources (above a minimum quality level) sounding the same.   The only theory that anyone could come up with (to explain these differences people are hearing) was that different sources must contribute different environmental effects (vibration, EMI, RFI), and that these must somehow be impacting the DAC's output stage.   Of course, using Toslink instead of copper should mitigate many of these effects.

The DAC white paper is worth a read if you are so inclined.

Hook
Posted on: 19 January 2011 by aysil
Hi Hook,
Actually, the DAC white paper was just open before me when I read your post! It displays very well actually how complicated the issues of precision and noise in digital processing are. Therefore I tend to perceive the claims of this buffering/re-clocking architecture not in absolute terms but in relativity; expressions like "eliminating ALL jitter" sound exaggerated to me.
Posted on: 19 January 2011 by js

Not if you think about it. There's more to this than jitter, like, how much influence do incoming abberations play on the jitter corrected result. Jitter free isn't the same a perfect but it appears the Naim technique of jitter correction is the current state of the art.

Posted on: 19 January 2011 by aysil
js,
thank you for reminding about possible mains interactions. I use Audio Magic power conditioner; all compared devices were connected to the digital section and powered at all times (except when changing connections). Digital, low level, and amplifier sections are isolated from each other, but to my knowledge no isolation among the outlets of the same group. Therefore, I cannot exclude the possibility of mains interaction. I will check, although I don't really reckon it played a significant role in the outcome. My priority was to have all equipment equally warmed up at all times. I experienced this is very important in comparisons.

About your question on video bits on the rest of the system: I don't even have a TV in my home!

You're right! Many factors play a role in evaluations. I see that you are selling Naim and other British equipment in the States. I guess for you GBP/USD exchange rate plays an important role
Posted on: 19 January 2011 by aysil
Reference:
Not if you think about it. There's more to this than jitter, like, how much influence do incoming abberations play on the jitter corrected result. Jitter free isn't the same a perfect but it appears the Naim technique of jitter correction is the current state of the art.

Exactly! This is what I tried to say. I thought jitter correction does not mean eliminating all kinds of error. Otherwise, I am well aware of the superiorities of nDAC.
Posted on: 20 January 2011 by realhifi
"I was not able to compare the UPnP output of the two devices with their S/PDIF outputs – I am waiting for the NDX! I also don’t know yet if the quality difference between their UPnP outputs will be as pronounced. I am pretty sure, though, there will be some difference."


Not sure what is meant here.  UPnP is a streamed network file is it not?  It certainly would be interesting to hear the differences between an nServe and HDX or for that matter a UPnP player such as Asset on a remote NAS or computer into an NDX.  I would asume that streaming UPnP into a renderer should level the playing field a bit? 

Now I'm beginning to see the reasons some are clammoring for a reference streamer that would have digital out to whatever dac was supplied.  Getting away from having the HD directly attached to the Dac
makes sense in a number of ways not withstanding the convience factor of simply not having the boxes in the listening room.

Now all we need to do is to get Apple to have some sort of streaming protocol for higher resolution networked streaming.  Unless that's already coming?
Posted on: 22 January 2011 by Mike Smiff
I can't ever see Apple building in high-res protocol unless they start to offer the high-res material for sale in the i-tunes store and lets face it i-tunes core is still aac and it seems that the mainstream apple/i-tunes customer is happy with it!

Sorry for the steer and roll on some NDX or UPnP thoughts, findings and views.

CD quality is fine but roll on 24 bit something or other mainstream and clasical music downloads.
Posted on: 09 April 2011 by aysil

When I had compared HDX and UServe and found big sonic differences btw these source devices through my reference dac at the beginning of this thread, many people asked me if I hear the same differences also through my nDAC. I could not answer this question for a long time, because of my travels and work load. Sorry about the delay!

Some had suggested nDAC would level out the difference because of its “buffering/re-clocking architecture”.  I doubted very much it would, and also doubted if it would be a good thing about nDAC if it would do that. Maybe, I thought, the extra processing the nDAC is doing to achieve this effect is the reason why I perceive the music from my previous dac (Kondo KSL-DAC) as more “pure”. This dac goes even without oversampling and digital filters; and maybe through this simpler architecture, it leaves the signal clearer and less stressed.

Those were just thoughts in my mind before doing the suggested auditioning. The results were not as expected; and deserve opening a new topic.

Btw, I updated my profile for those who were curious about my system.