The new Naim DAC

quote:

Originally posted by Adam Meredith:

quote:

Originally posted by connon price:
Some differences could also arise from difference between Coax and Optical transmission.

In fairness - the Apple doesn't "sound" with co-ax (I am lead to believe).

Sorry if I didn't make that clear. Yes, Apple is TOS and HDX is Coax, so even if they were both playing the same high quality (HDX?) rip and are comparing apples to apples, they are being squeezed through different presses on the way to make apple juice. A difference could (would) result.

hm...

Why not a BNC connection?

Generally, the voicing on the coax cables varies greatly whereas for the optical's tonal balance was very similar. ( quality varies plastic vs. glass optical and their terminations )

I found Naim DC1 to have warmish and has a good body over the optical cables I have tried in the past.

quote:

Originally posted by kuma:
hm...

Why not a BNC connection?

There are now 75 ohm spec RCA connectors thanks to video so things are no longer so dire but I also like BNC. I believe they are simply following the appropriate standard.

Pity.

They ( Naim ) do not want to be accused of *funny cable* brand again, probably, too.

BNC connection never took off in the market place, oddly.

I'm very glad we are finally getting one of these. Most DAC's are pretty small. Why is this new Naim DAC in such a large box? Adam?

quote:

Originally posted by Edot:
I'm very glad we are finally getting one of these. Most DAC's are pretty small. Why is this new Naim DAC in such a large box? Adam?

Big Box, Big sound.

Funny you asked as many others have said the box wasn't big enough.

quote:

Originally posted by js:
I thought about it on the drive in and couldn't think of why transformer was required for an AES/EBU output to Spdif in and wondered if it could be done with resistors off of one leg so I did some checking. While a transformer will be needed for a balanced dig in from single ended out the reverse may not be true. Here's a link to differences in format and how to match by building resistors into an end plug. http://www.epanorama.net/documents/audio/spdif.html Looks like pin 3 is shorted with impedence matching and level reduction done with a resistor bridge to ground off pin 2. I would keep the cable under 2M in this configuration and frankly, I think they almost all get worse over 1M anyway.

Canare, Neutrik and Graham Pattern Systems make xformer based adapters in either direction.

AES/EBU to S/PDIF signal level converter

AES out:
2+------330 ohm------------------- SPDIF in
.............................|
3+.....................91 ohm
|............................|
1---------------------------------
|
-ground

Here you go for the diagram. Ignore the period dots and you've got it. Verticle lines show connections. Needed to use characters of some sort to position things. If it were me, I'd short pin 3 at the XLR end and built the resistors into the RCA.

quote:

Originally posted by js:

quote:

Originally posted by kuma:
hm...

Why not a BNC connection?

There are now 75 ohm spec RCA connectors thanks to video so things are no longer so dire but I also like BNC. I believe they are simply following the appropriate standard.

BNC is my favourite connector - I hate RJ45 with a passion: horrible tacky plastic things.

It was regarding connection to and what was available on the DAC.

sorry, no but feel free to email. I'm typing handicapped anyway so you're better off. js95@sbcglobal.net

Looking at that diagram I copied from the site, it seems to me that the values could be more optimized. Perhaps it's the best compromise and a xformer is still preferred but it's worth a try for those that want to give it a shot as it may be close enough in impedance to not matter much.

No I didn't know skype though I knew those sort of things existed.

quote:

Originally posted by js:

quote:

Originally posted by goldfinch:
Js,
No I haven't tried any adaptor, I just ordered a suitable cable (Belden AES/EBU) for connecting a Lynx AES16 to a Lavry. I think you know I was having sound dropouts with via Spdif caused by mains, but they disappeared since I changed to an AES/EBU interface. At least I thought the stronger signal of the AES/EBU interface was the solution to that problem.
On the other hand, I think some of the AES/EBU advantages you point out can be useful in domestic environments. For instance, long cable runs (you can keep the computer far from the system), it is more interference-proof, this was exactly my experience, and finally, user could choose among more audio devices for streaming (pro and not so pro, I think most new music servers-streamers have AES/EBU).
So, it would be a pity that Naim brand new DAC wouldn't have an AES/EBU digital input.

The longer cable runs still suffer from from the same bandwidth and cable character issues as any single ended cable. Common mode noise cancelation is the only advantage and you hear the same differences in cable type and length as you do with Spdif. The dig files are stored as single ended files. A second out of phase signal needs to generated somewhere in the chain which to me is just complication though not a problem if done well. Generally at the outs and ins. I could be wrong but I believe the Spdif that you were using was TOS and direct fron the computer. Though it wouldn't have mattered in your case if it were tos or coax for this discussion I'm only referreing to coax. The Lynx works because it's simply a better card with well sorted buffers that can also be increased if dropouts occur. It's the card and not the interface. You can run a coax 30' though I wouldn't do that with any dig format I was trying to get the best from. AES/EBU does have a significantly greater distance spec but that doesn't make it better for getting the ultimate performance from the format. Just more managable for pro use that sometimes requires AES/EBU to get a job done at all. Being able to do something and doing it well aren't the same. There's no advantage at normal lengths. If streamers have AES/EBU it will be at the analog out to cater to those with balanced preamps. Not the DIG out.

Hi Js, I did a little experiment in order to test whether an AES/EBU connection helps or not to avoid the sound dropouts I was experiencing.
I have tried with different coax. cables (cheap ones and chord signature) using different sources:
(A) standard spdif from cheap computer sound device integrated in the motherboard
(B) Spdif from an M-audio 192 audiophile sound card
(C) Spdif from a fine Denon DVD 3930, which is supposed to have a very good quality digital out.
(D) Spdif from a cheap pioneer DVD.

All were connected to the Lavry's Spdif input and in all cases sound dropout happens when I turn on/off some lights at home. Cheap or expensive digital cable didn't change anything.

With the AES/EBU connection between the Lynx sound card and the Lavry there isn't any sound dropout.

Couldn't be that the stronger voltage of the AES/EBU connection make the digital signal more insensitive to mains interferences?

I am sure Naim DAC will provide the most complet array of digital inputs, not only AES/EBU but even I2S if this is really a far superior connection...

Other good point with AES/EBU is the XLR connector, I think this is better than any other, you can hear a "click" and you know the connection is flawless...

What do you think guys about what appeared in WHF?:

It's very much work in progress, but is being developed as "an innovative way of turning digits into analogue", according to a Naim spokesman. "It's designed for people who use a laptop or PC to play their music."


What can mean "innovative"?

- A proprietary conversion mode: We just know only a litte about how the SN DAC works and deal with jitter, how different can be the new DAC design?

- Convenience features still to be implemented in the prototype suchs as wifi (like new Chord DAc) or built in sound device for attach a NAS.

The fact it's designed for "people who use a laptop or PC to play their music" sounds to me as a simple DAC to connect to the Mac and that's all, I mean no I2S or AES/EBU connection or other features more common in the pro audio DACs as polarity, volume, different filters or clock modes, external clock upgrade, headphones out).

I think a 2000 Naim DAC upgradeable with something like a 555PS should be "pro audio" friendly, at least my wishlist features would be:

- Complete array of digital inputs including AES/EBU
- A conversion mode sensitive to input signal quality. I mean that if the transport used is top the DAC conversion should take advantage of that and be able to give the best results but at the same time if the digital input is, for instance, full of jitter the DAC should be able to cope successfully with it and give a good result too.
- Full HD res. input compatible, I think USB is limited to 24/96 though,
- NAS or Wifi compatible (optional?).
- "S" grade performance potential with all the fun you can expect from Naim: rhythm, dynamics, lots or PRAT...

What do you think?

Personally I'll be happy if it has 4 coax inputs, since they are still the most reliable and widely-usable/compatible. It's bound to have others too but 4 coax should be the start.

quote:

Originally posted by rupert bear:
Personally I'll be happy if it has 4 coax inputs, since they are still the most reliable and widely-usable/compatible. It's bound to have others too but 4 coax should be the start.

With a standard length box there is a lot of space...
Inputs
4 Spdif (maybe one of them with BNC connector)
1 optical
1 AES/EBU
USB
Ethernet

Burndy PSU socket
IEC mains

Outputs
Din
RCA

quote:

Originally posted by goldfinch:

With the AES/EBU connection between the Lynx sound card and the Lavry there isn't any sound dropout.

As I know, but I could be wrong, there is a small transformer on the AES input of the Lavry and this can break some ground loop somehow. It can cause a behaviour like this.

Too bad no one took a higher resolution photo, or a photo from behind... Still, since they do say it's still "in development", I guess it doesn't matter.

Just in case anyone from the development team is reading:

- We do need a cheap (2K is an OK price) DAC. It doesn't have to beat a HDX, and even the option to upgrade is great. It should however, have a computer connection. USB is OK, but I do think that a network connection would work much better. Having digital inputs as well (S/PDIF) is great, and adds to functionality, but I would be much happier with a network port. Improving other devices is fine, but a proper source is essential. Since Naim already has the software developed, for HDX, it shouldn't be difficult, or very expensive.

- Another device, the often mentioned HD 555 would be nice as well. Even a better DAC then demoed, the works, and in an another price range... But that's out of my league, so I'll skip the comments.

Bottom line: DAC is nice, but do add a network port as well. Pretty please?

Can't wait for May, to see the final specs.

quote:

Originally posted by ferenc:

quote:

Originally posted by goldfinch:

With the AES/EBU connection between the Lynx sound card and the Lavry there isn't any sound dropout.

As I know, but I could be wrong, there is a small transformer on the AES input of the Lavry and this can break some ground loop somehow. It can cause a behaviour like this.

Thanks, I didn't know which method they used but this is common. An extra bit in line to allow balanced. Can easily be done outboard. The xformer does not isolate the groung in this configuration and if your having a problem with Spdif, you need to look at the hardware, not the format. No advatage over Spdif in normal lehgths. If you like it better, it's because of your source card quality or you happen to prefer the 'character' of the input which is OK too. The voltage spec will have nothing to do with it under about 10M.

More than one optical in, pretty please

Not sure if these have been mentioned earlier:

The prototype shown was very much pre-production. Anything can change - case, inputs, outputs, electronics, PCB mounting, power supply options, internal configuration, filter settings. All still up for grabs.

The case was chosen for reasons of cost. A Reference case would raise the price significantly, whether Hicap or full size.

It won't stream from a NAS or anything.

8 inputs was mentioned, but no suggestion as to their configuration - i.e. whether they'd be AES/EBU, coax (RCA or BNC?), optical.

No mention of digital outputs that I heard.

I was concerned about jitter and asked about that. Apparently, Naim have devised a novel way of dealing with jitter and clocking generally. This work came from the DSP work that was done in the Bentley project. If it works half as well as we heard on Sunday, it's possible it'd make the Bentley project worthwhile on its own! I believe a white paper may come out in future discussing this approach to managing jitter.

My only wish would be ASIO capability for the USB. Would be good for both pro or advanced home use. I can get around that but it would simplify things.

quote:

Originally posted by Frank Abela:
Apparently, Naim have devised a novel way of dealing with jitter and clocking generally.

It would be very, very interesting to hear more about this. It is my understanding that NAIM have always tried to keep the DAC close to the source of the signal, presumably to reduce the effects of "jitter" in all of its forms. For the first time, apart from the DAC in the SuperNait, NAIM are releasing a DAC distant from the signal. Given the history of the company and, one would guess, the internal culture it is far from surprising that NAIM, from this statement, have put time into a novel way of either preserving the signal, recreating the "true" signal or some other electronic loveliness to present the truest signal possible to the actual DAC.

Is there anyone who can talk a little about this? I would guess no: but one can but ask politely

I'm also VERY interested in hearing about how jitter can be managed as it's my only concern when bolting on an external DAC. It's most commonly dealt with by buffering and reclocking within the DAC, but with DSP you can perform many complex tricks.

No idea what they're doing but look at white papers about Dice used in the TC or JET in the Lynx. Having a direct asynchronus usb computer connection with that sort of clock attachment inside a really good DAC would be monumental. Even when using those excellent devices DIG outs, you can still get small amounts of jitter on the cable, transmitter and receiver. Their PS's are also a source of jitter on the way out or even before and the clock needs syncing. Still better than the reclocking methods I've come across when the jitter is reasonable but who knows, maybe they've discovered a way to build a better mouse trap. Won't know until a listen.