A DAC, is a DAC, is a DAC......discuss

"The change of color is likely and a difference a very little difference is prepared. Sugar is not a vegetable."

dac design hasnt really changed for ages. manufacturers have learnt to use them with good pus. the market is crowded  now the competion begins  arcam started it with the black box they are saying their new one is very good

Geoff, interesting comments.

I am listening to Ayon Network Player for the last couple of weeks; and I like it much more than the KDS I had demoed some time ago. I hope to make a direct comparison soon. You imply in your post that the more significant differences will be in the implementation of the analog stages.  You may be right here in that the tube output stage of Ayon may more be the deciding factor for my liking, to my taste.

However, when I look into the design approaches of different DACs, I still see somewhat big differences. Just an example, implementation of a buffer before the dac chip, and even to use or not to use a buffer, seems still highly controversial. Therefore, I don't expect a "leveling out" of dac designs any time soon. I would just agree if you say we should not neclect the analogue components of our system because of distraction onto the current competition in digital devices.

Thanks for your interesting comments guys.

On the software I do agree this is the real tough bit for all manufacturers. Naim and Linn are examples at opposite ends of the problem.

Linn's open source approach has its merits and they cleverly ( or maybe not so cleverly?) abdicated responsibility for the ripping/creation of the music library and data delivery to the streamer/DAC...which pushes a significant responsibility onto the dealer and /OR the end user. That can be quite a rocky passage for the newcomer to this whole mare's nest.

Naim started by taking on the whole thing and making it user-friendly in their early boxes which had CD ripping and HDDs and internal software to read the HDD and play music. Unfortunately as they evolved into NAS and network delivery of data outside their boxes the software to deal with this needed to get developed and they had made it difficult by starting out owning the whole process internally.

I had a love affair with Naim CDP's and still am likely to get dewy eyed when I think off the sound CDS3 / 555PS made whatever its absolute shortcomings in terms of purity over coloration might have been in retrospect. In fact I think the coloration etc of CDPs will be a loss as streamerDACs gain in the market. There is a danger that the potential purity of their replay chain can turn music more analytical more quickly unless the people voicing the DAC products have a grounding in what constitutes enjoyable replay. A friend recently complained that CD555 replay made his odd crap poorly mastered CDs sound acceptable whereas the ripping/ streaming/DAC process was so unforgiving they sounded as bad as they really were. I guess a good CDP can 'polish a turd' better than a DAC can.

Aysil your points about the implementation are interesting. The size of buffer I think is a function of how the data gets to the DAC, but fundamentally once in the DAC circuit processing differences are minimal to final SQ, Again in following analog stages there is a lot more scope for variation but this is typically user determined since I think most folk use an outside pre-amp with the more advanced DAC products. On the other hand maybe by keeping the signal paths short and internal, products like the Uniquite benefit out of proportion to their price level.

regards

Geoff

AllenB posted

I totally agree that implementation of DAC circuits is not new to our beloved manufacturers, I can remember arguing quite fervently when the HDX was released (remember this was before the nDAC) that Naim had all those years of CDP development logged and they should know how to make the best of internal DAC's. Why couldn't Naim have taken the basic engineering in a CD555 and replace the CD mechanism with a network streaming board. I'm sure I am seriously over-simplifying things, but you get my gist. It seems like an age for Naim to be in a position when the CD555 might be matched or bettered, and if that is borne out, we should all be ecstatic, after all the CD555 was / is the very best exponent of the CD medium. Will it ever be improved upon? Not likely on commercial reasons, and also probably not possible from a SQ perspective.

That Naim did not lead the field from the off has horrible parallels with Kodak refusing to take the lead with digital photography once they had invented it, in case it dented film sales.

Naim really does need to assess the future and start to catch up. I am sure that the CDS3 was discontinued because it no longer sold, and the same fate cannot be far off for the CD 555, when other makers can get the kind of quality it is offering  without a CD drive involved in playback.

ATB from George

Hi DACs vary from implementation  to implementation.

Leaving asidemechanical, decoupling, earth currenys, powersupply regulation and filtering, EMI management, which all have a part to play, not least because we have some high energyclocked digital circuitry in close proximity to sensitive analogue circuitry.

But moving into the heart of the matter, the DAC can onlbe ever be provide an approximation in the reconstruction of. Ananalogue signal. Goodoils maths, physics and electrical engineering show why thiI can only ever be. Therefore in the process of reconstructing the analogue signal with widest bandwidth, compromises and choices are made by the designer. Ie to focus on some attributes at the expense of fothers. These compromises vary from design to design, and include DSP, converter chip, transform filter function, analogue filter and linearity of reconstructed analogue current. 

These all produce a 'sonic character' to a DAC.

Now although much of the DSP maths has beenaround for 20 years or so, advancements are still made in this area, but the key thing is the ability to more accurately digitally process the signal to aid transformation. Here recent advances in fast floating point arithmetic signal processors has helped.

Therefore DACs vary quite considerably from design to design.

Simon

Hi Hook - I'd prefer Naim to select the best filter option rather than offer user selectable options. Most are fairly subtle and as you say users have a play and then settle with one. My Weiss had selectable filtering but after an initial play, I settled on one and forgot about it. 

Power supplies - beyond more work on reducing noise on the individual rails, there is scope for increased isolation between the various sections. More in curcuit regulation and if the NDS uses the 555 PS more scope to use the multiple power supply rails on offer from this supply. As the NDS is designed from the outset to use an external supply, the internal wiring routes can be optimised to the burndy connectors. Given the extra real estate available within the casework, internal layout can move  very noisy sections - streamer board and associated components - away from the more sensitive DAC and Analogue stages.  I look forward to looking under the bonnet of this one 

James

Hook, I agree, offering filter options would be good especially on revealing systems  albeit possibly very complicated to correlate digital and analogue filtering and do it well. But as there as no such thing as a 'best' filter this would be really usefuhost as to better match digitalcharacteristics of the source and personal and downstreamm system preference, and with a Naim system I would like to see the filter optionencoded as metadata so it can be assigned to particular content. Now that would be neat :-)

Simon

Simon's informative postings confirm there's rather more to DACs than might be apparent to us less technical bods, yet I think one of the surprising things is that the wildly differing implementations can result in remarkably subtle differences for the listener.

Having heard a CD player that betters my 555 I'm certain there's more to be extracted from the medium but   it's unlikely anyone will make the investment at this stage. Perhaps in a few years we'll see a "CD Revival", much in the way that the LP was all but dead and buried for a while yet has staged a remarkable recovery recently.

Anyway, I think there's considerably more to come around the development of DACs and their implementation. Separate power supplies work for Naim but bring their own problems. Linn in their Klimax seem to have managed to incorporate a high-quality switch-mode power supply into their box of tricks for those who want a one-box solution. Then there's the business of getting the digital information from source into the DAC itself. I reckon there's lots more to come yet.

Tony, I agree there will always be improvements or differences,  other than with my current CD purchase rate, I don't want any revivals just yet :-)

Although  the top end Linn streamer sound just doesn't seem to do it for me there is nothing wrong with SMPS, there is no reason why they can't be engineered very well down stream. iME it's often the upstream area where SMPS can cause issues. So in my book, I feel you must either go all switched mode including amps, or try and decouple as much as possible between transformer and SMPS units.

In some of my other areas, I deal with SMPS, including some very sophisticated ones, and they do tend to produce 'birdies' or if very precisely clocked, spot harmonics across the spectrum on the mains, and this does tend to decouple back into sensitive audio circuits as intermod if you are not festiduous with cable dressing, routing RF filtering,  and spacing. So in that sphere the choice is to go linear where possible albeit at great weight and cost.

Simon

I do wonder how good is good enough.

When the point of reference is exceeding vinyl replay for musical satisfaction - a medium capable of excellent quality, but ultimately flawed, i wonder whether the best in digital will ever be 'good enough' for some. 

Tony - which CD player bettered your 555 ?

James

James, it is worth noting vinyl is master cut using hidef digital audio to the cutting head electronics in this case the cutting head forms  part of the digital to analogue transfer function....

Hi James. It's a rather famous prototype owned by a Forum member and heard by a few other honoured Naimites. Respecting the wishes of the owner I won't give any more details without his permission.

Tony - ah yes. I know the one you mean.

Simon - yes good point. Digital intrudes into many analogue domains. The wonders of radio 3 on FM spring to mind.

James

True for new issue stuff. The original issue Vinyl dating back to the 50's and into the 60's had to make do with an all analog chain 

...and damn good a lot of it is

Geoff

Interesting comments. thanks folks.

Does anybody know if this technique of paralleling multiple DAC's that some are touting has any genuine benefit?

It seems to suggest one pass through a single channel DAC isn't good enough to extract anything thats there which I don't see.

Geoff

With vinyl resurgence at the moment I wonder why mastering studios do not revert back to analogue master tapes so we could enjoy modern AAA recordings. This wouldn't prevent AAD or ADD to please digital nerds.

Cat - Economics - The vinyl resurgence is small beer compared to the mass market digital demand. I also suspect that most producers are rather better versed these days in using pro tools than a big analogue desk and a couple of Studers. Shelby Lynne is one artist well known for her love of recording to analogue tape and taking care in the subsequent production and the end results are sublime. 

Noogle - yes the distribution. 

James

James, I see that you have gone from computer audio to different streamers. From your point of view, which is the best solution if one would like to retrieve the best sound quality from internet radio. Is this more a matter of the software used to tune into different stations or does it have more to do with the hardware used? 

I've never really delved into the depths of Internet radio although from a usability point of view the NDX made it very easy to tune around the various 'stations' on offer. 

James