Naim DAC and USB memory stick compared to Naim DAC and CDX 2-2

quote:

Every time I hear that different USB sticks sound different, or different cables sound different, or different NAIM transports sound different through the DAC I become less impressed with or interested in the DAC.

licemusic, I understand your frustration because dishonest advs can ruin the hopes and eventually lead to losing clients. Either Naim was not very precise or some other people misinterpreted the original advs or was it an preheating (or overheating?) of the delayed DAC release -- but somehow the idea of total jitter rejection in the forthcoming Naim DAC has penetrated to internet. It's worth mentioning that from theoretical point of view S/PDIF is definitely not the best way to conduct non-jittered signal to the DAC kernel. It also worth mentioning than Naim is not the first manufacturer who built a re-clocking DAC. Why should we think they will do what the other failed using quite conventional concepts (here I'm talking of the digital path only)?

I think we should not wait for ridicuolos features from Naim DAC. If a bitstream is irreversibly spoiled by jitter -- no DAC in the world will repair it. Take Scarlatti as example -- they paid crazy attention to the jitter and clock issues. But it will not make DVD-player sound the same as CDS3.

Meanwhile the modern hi-fi is developed enough to offer high quality digital transports for affordable price -- I would say less than 3 K$. This will develop further within computer-based systems. I believe CD5XS and CDX2.2 are intermediate steps on this path and Naim will quite possibly come up with dedicated CD-transport for circa 1.5 K$ with the same jitter quality as CDX2.2 and many chances they will build a pure streamer without HDD and DAC.
If they don't -- Naim users will pair Naim DAC with no-Naim transports instead of buying more expensive all-Naim solution based on CDX2.2. Nobody wants to pay twice for buying expensive and unnecessary staff.

Buying no-naim transport is not a big a deal by the way -- except using two remotes instead of one.

Also I don't think USB flash can make a difference on the sound unless it's malfunction and leaks a signal to the ground. But digital cable and different digital sources do have an impact on the DACs. Of course, reclocking DACs improve the bitstream more than PLL-DACs.

AMA , the Linn DS solution is increasingly looking a lot simpler, and more convincing. Linn are quite clear that in their DS system cables don't make any difference, ripping formats don't make any difference and that their products have the best power supply possible inside them, rather than available as an expensive extra. In their solution, musical data is kept in the digital domain guaranteed bit-perfect until it is clocked out of the buffer into the DAC - no s/pdif nonsense, or bizarre half-assed USB stuff, or trusting to a mish-mash of media-players, sound-cards and device handlers which may or may not be bit-perfect even before the interconnects get the data.

quote:

Originally posted by likesmusic:
AMA , the Linn DS solution is increasingly looking a lot simpler, and more convincing. Linn are quite clear that in their DS system cables don't make any difference, ripping formats don't make any difference and that their products have the best power supply possible inside them, rather than available as an expensive extra. In their solution, musical data is kept in the digital domain guaranteed bit-perfect until it is clocked out of the buffer into the DAC - no s/pdif nonsense, or bizarre half-assed USB stuff, or trusting to a mish-mash of media-players, sound-cards and device handlers which may or may not be bit-perfect even before the interconnects get the data.

Likesmusic
You say Linn says cables don't make any difference, but have you listened to CAT5 Vs CAT6. There will be different RF reflections and couplings changing system noise.

Also you say Linn say PSUs make no difference, but they have just upgraded their PSUs. It's just Linn don't offer an external one.

You say Linn says ripping formats don't make a difference when quiet clearly they do.
Obviously MP3 Vs WAV, but even FLAC can sound different due to the significant extra load on the decoder.

You say UPnP streamers are bit perfect and imply S/PDIF is not. S/PDIF is the original 100% every time bit perfect transport. No messing around needed. Check out different UPnP servers, not all are bit perfect, some are very bad giving only about 14 bit resolution!

An S/PDIF DAC is the only way to create a simple digital interface that is bit perfect.
All other methods need special care including UPnP.

Likesmusic, how do the linn streamers output ( into the preamp ), no s/pdif nonsense obviously!

quote:

Originally posted by 110dB:

An S/PDIF DAC is the only way to create a simple digital interface that is bit perfect.
.

Do you connect your printer to your network via s/pdif? Are you saying that ethernet protocols aren't bit perfect, or simple?

Granted that s/pdif can be bit perfect; but that doesn't mean it is electrically perfect, otherwise why would there be so many posts on here saying that even different Naim transports sound different into the Naim DAC?

Checkout the Linn forum; there are many posts from Linn design engineers saying that in a DS system CAT6 sounds no different from CAT5, and FLAC no different from WAV. They have evidently designed their products to be uninfluenced by the different RF reflections or extra decoder loads of which you speak.

licemusic, you are absolutely true in terms of logistic and technical values.
But there is one more variable in all these considerations -- commercial.
I remember Stewart Tyler (ProAc) said that he can build absolutely neutral speakers, perfect from technical viewpoint -- but nobody will buy them. Many people over the world use different digital sources -- CD transports, DVD transports, streamers, PCs, Macs. Linn DS cut this down to a very narrow range of NAS-based Twonky-drived audio replay. If they picked up the strategy correctly -- they will not lose. I personally believe this approach is strategically WINNING. But at the moment Naim DAC is much more versatile solution than Klimax and is distributed in three configurations: bare, XPSed and 555PSed which makes it much more affordable to start with and possibly stronger (in Naim values) than Klimax when paired with 555PS. I expect that Naim DAC will be well received in non-Naim community and possibly collect a lot of awards and bring more income to Naim than Klimax to Linn -- while being a step behind the Linn DS ideology!

But in time -- if market defines a streaming as a leading trend then Naim will definitely design a Naim DS -- by removing unnecessary interfaces, removing built-in PS, springing the board, placing more regulators on board and .. increasing the sale price for the new device

My advice is to follow the market, buy Naim DAC and if the new device will come -- upgrade for the new one. Changing to Linn is not a solution for those who value a tight sound.
I'm standing in a queue since October and don't have even a smallest hint on when this black box will be shipped... I think Naim commercials don't know themselves when they gonna provide the next shipment.

quote:

Originally posted by DHT:
Likesmusic, how do the linn streamers output ( into the preamp ), no s/pdif nonsense obviously!

They have analogue audio outputs - the DAC is inside. True, the lower end models have s/pdif outputs too, but I doubt they are much used.

quote:

Originally posted by likesmusic:

quote:

Originally posted by 110dB:

An S/PDIF DAC is the only way to create a simple digital interface that is bit perfect.
.

Do you connect your printer to your network via s/pdif? Are you saying that ethernet protocols aren't bit perfect, or simple?

Granted that s/pdif can be bit perfect; but that doesn't mean it is electrically perfect, otherwise why would there be so many posts on here saying that even different Naim transports sound different into the Naim DAC?

Checkout the Linn forum; there are many posts from Linn design engineers saying that in a DS system CAT6 sounds no different from CAT5, and FLAC no different from WAV. They have evidently designed their products to be uninfluenced by the different RF reflections or extra decoder loads of which you speak.

Likesmusic

Two hypothetical things:

No, I don’t connect my printer via S/PDIF, but if there was an S/PDIF interface it would be word perfect.

I wouldn’t connect my printer via a UPnP server without fully testing it as it could put the wrong letters and words in! – fact!

Now real stuff:
No, I was not saying Ethernet protocols are lossy, I was stating the fact some UPnP servers are lossy. I was also stating the fact S/PDIF is bit perfect and always has been.

I was also stating the fact all electrical high speed electronics are subject to RF reflections. For example on a broadcast transmitter if you get this wrong you wave ‘bye bye’ to your output circuits.

I’m sure the Linn engineers say CAT5 is perfect. Philips engineers said CD was ‘perfect sound for ever’ too. Would you run your CAT5 cable near your power amp input cables? – no? Because you know it would sound different. What would happen if you changed the twist ratio in the CAT5 cable like CAT6? Would that interfere differently? Yes it would.

quote:

Originally posted by 110dB:
I was stating the fact some UPnP servers are lossy.

So what? Use a server that isn't lossy, such as Twonky or Asset. Or an in-house solution.

quote:

Originally posted by 110dB:
I was also stating the fact S/PDIF is bit perfect and always has been.

So what? Why does the Naim DAC sound different with different Naim transports?

quote:

Originally posted by 110dB:
I was also stating the fact all electrical high speed electronics are subject to RF reflections. For example on a broadcast transmitter if you get this wrong you wave ‘bye bye’ to your output circuits.

So what? You just have to get it right. It is called engineering. Downloads manage to get to your pc through dozens of computers, servers, cables, telephone exchanges and who knows what else. Digital data is remarkably resilient. Why on earth do you want to mix it up with a clock signal to get it into a DAC which promptly discards the clock, but nonetheless seems to suffer from noise introduced by that connection?

quote:

Originally posted by likesmusic:
Downloads manage to get to your pc through dozens of computers, servers, cables, telephone exchanges and who knows what else. Digital data is remarkably resilient.

Though this is obviously true, the quality of the data isn't time dependent to the same extent of real-time audio or video.

Protocols are designed in very different ways to accommodate the business requirement. I.e. TCP/IP for computer networking since data packets are lost constantly, and will need to be re-transmitted so you receive an exact copy. Standard UDP on the other hand drops the packets, so this can be used for real-time market price information, since it’s constantly changing, and therefore you do not need to receive an exact copy of a previous price quote, which could lead to data delays (buffering) highly dependent upon the receivers and senders bandwidth. In effect, digital data is not resilient, unless protocols, error correction, or other maths are implemented to the meet a satisfactory level of performance.

It really isn’t that simple, but it isn’t snake oil either.

quote:

Originally posted by js:
FLAC, neat trick. Don't want to give the wrong impression.

Sorry - should of clarified, I won't be needing a DAC to check the differences of our USB sticks, the differences are still clear on the Uniti and the NS01 server we currently have in stock. Certainly, in respect of 24bit FLAC, we could not use the DAC or uniti anyway....

quote:

Originally posted by PMR:
In effect, digital data is not resilient, unless protocols, error correction, or other maths are implemented to the meet a satisfactory level of performance.

Are you saying that data is lost or corrupted when I download a product from the Naim music store? If you are then hopefully you can demonstrate it for it follows the store is a waste of time. If you aren't then surely the transmission medium is adequately resilient. Of course protocols and error correction etc. need to be implemented; and they are; and they work.

quote:

Originally posted by likesmusic:

quote:

Originally posted by PMR:
In effect, digital data is not resilient, unless protocols, error correction, or other maths are implemented to the meet a satisfactory level of performance.

Are you saying that data is lost or corrupted when I download a product from the Naim music store? If you are then hopefully you can demonstrate it for it follows the store is a waste of time. If you aren't then surely the transmission medium is adequately resilient. Of course protocols and error correction etc. need to be implemented; and they are; and they work.

Yes, I agree. TCP/IP ensures the file download is perfect or it fails. In the same way that a data file can be copied perfectly from a data CD unless the error correction fails. However, the time it is copied various enormously, so you need a perfect buffer and clock to ensure no audible differences with an audio CD.

quote:

Originally posted by likesmusic:
So what? You just have to get it right. It is called engineering. Downloads manage to get to your pc through dozens of computers, servers, cables, telephone exchanges and who knows what else. Digital data is remarkably resilient. Why on earth do you want to mix it up with a clock signal to get it into a DAC which promptly discards the clock, but nonetheless seems to suffer from noise introduced by that connection?

Despite all our computer geeks "common sensical" understanding of how data is transmitted and treat music as just one form of data, the Holy Grail counter-argument used to be jitter. And if jitter cannot be part of the equation (e.g. async data), then there is always the interference “problem”

Without “problems” like the above, there cannot be “solutions” like audiophile CAT5 and USB cables. Fortunately (or unfortunately depending where you are coming from), using USB sticks as audiophile music playback is not common practice, or we will see $5000 USB sticks.

We audiophiles are very vulnerable to the slightest power of suggestion; we are all guilty of that in one form or another being in this forum. That’s part of what makes this whole hobby fun.

But how far should we go? Should I go and check my WAN router to make sure that it is downloading music from HDTRACKS audiohpilely? Should I put spikes under my LAN switch?