Weiss 202 - a few thoughts.

Posted by: james n on 26 July 2010

Ok i though it was time to put fingers to keyboard now that i've had a few days to play with the 202.

My DAC history up until now has been the n-Vi internal DAC, The Lavry DA10 and then the Weiss DAC2. I'd seen a review in Hi-Fi News of the Weiss Minerva (the DAC2 in fancy clothes) which got me interested. A quick chat to Keith at Purite got me a loan of a DAC2. My DA10 went to its new owner (who now has a DAC2) and the DAC 2 took it's place.

The Weiss via optical from the Mac was not that different to the DA10 and I could have quite happily gone on living with the DA10 which to me combined all the virtues of a Naim sounding front end married to the convenience of iTunes. Via Firewre it was a whole different ball game the Weiss clearly better than the DA10. Amarra and Pure music have been icing on an already rather nice cake and up until a few weeks back i'd arrived at a rather fine sounding front end.

Enter the 202. I'd not been too interested in this thinking it was just a minor update of the DAC 2 with the addition of a nice front panel display and remote control. Not much benefit to me i thought. A Google around found some internal shots - the DICE Firwire implementation is the same but from there on - Twin transformers for more isolation of the digital section and a completely new DAC / Analogue stage using the Sabre ES9018 DAC chip. The 202 like the 2 / Minerva keeps the DAC section in a screened 'quiet room' - good to see in a device with a lot of high speed digital noise flying around inside its box. The 202 has two built in, user selectable filter settings, A and B. Filter A has a steeper frequency response than B and more filter settings will be added in future software updates. The other nice feature is the transparency checker which uses the playback of test files to check out whether the playback chain from the test file stored on the computer to the the DAC is bit transparent (of which Amarra and PM are when volume is set at 0dB)

Last year i popped over to Joe Bibb's to have a listen to his DAC2 / Berning / Art Emotion Signature system and we had an enjoyble day listening to his system - its a fine system and it also changed my mind with regard to valves. Like me, Joe had used the DA10 and moved onto the DAC2 so i found we had pretty similar tastes. I know if he's discovered something then its worth giving it a try. When Joe got his 202 and mailed to tell me how good it was i thought i'd better get one and have a listen.

The sound ... if you like the DAC2 then you'll love this. It just takes all the good things of the DAC2 and builds on them. A number of things really stand out on this DAC on familiar music. The sheer clarity for a start - all sorts of minor inflections in a voice are just so clear, cymbals just shimmer and have a very natural sheen and minor instruments way down deep in a mix are clear and easy to follow. I've also noticed that vocals that sometimes could be indistinct are now crystal clear and just sound so real. Close miked Cash (American recordings) or Shelby Lynne sounds superb. Tracks that may have originally have sounded harsh are now clear and detailed - its not a bright sounding DAC. The other thing of note is just how quiet the 202 is - it has an extremely low noise floor which translates to an inky blackness which particularly noticeable when a track is coming to an end and is slowly faded down - musical information is still there even though it can be just a whisper right up until the track stops. It just drags up low level detail i've not heard before.

You can probably tell i'm pretty pleased with my new purchase and if you already have a DAC2 and like the Weiss way of doing things then it's well worth a listen. The DAC2 is still a very fine DAC and with the Fiewire connection to the Mac its a neat plug and play solution.

How does the 202 compate to the Naim DAC ? - to be honest I dont know as i've not heard the Naim DAC for a while now but i'm sure given the level of engineering that's gone into the Naim design its a very fine DAC and in full battle configuration with the 555PS i'm sure there would be little to choose from between the Naim and the Weiss - its down to personal preference / system matching in the end.

Hope that's of interest to some

James

Posted on: 30 July 2010 by js

quote:
Originally posted by DHT:
When Naim bring out their async usb or async firewire dac ,no doubt the 'trade' members will perfrom a spectacular volte face.

How would "I've always liked it as a computer interface." when speaking of jet create a situation that would cause 'a spectacular volte face'. I was using firewire JET tech before you were.

I just understand that there are other viable options. Our best input into the Ndac is not the INT202 though it is the best from a direct computer hookup.

Wired Ethernet has very high capabilities and SPdif also if the stream is good. Hows your hard drive attached? Asynchronusly?

How could PCI not be the best for a computer other than system noise? I still hear significant differences from an INT202 when I change the buffers in the device, the cable or the power supply used. Nothing's perfect.

Posted on: 30 July 2010 by pcstockton

quote:
Originally posted by tonym:
There's no doubt even Firewire is a far less satisfactory means of digital transmission than 12S but the latter won't work over a distance greater than a few cm.

A few CM? The i2s fitted DAC, i2s mods, and USB>i2s converters from the likes of Empirical Audio, Stello, and Northstar severely disagree with that.

Posted on: 30 July 2010 by james n

i2s really is designed to chip to chip. Adding in some differential drivers allows it to go longer distances but its probably not the best way to do it without a lot of care taken in the implementation.

Posted on: 30 July 2010 by js

I've heard plenty of I2S and admittedly some combos were better with the option engaged but ultimately I agree with James n here. Best in proximity. I didn't think the SPdif reciever sections very good on their own in those dacs anyway so....

Even in async, the clock is in the device but it's controlling something not local.

Posted on: 30 July 2010 by AMA

Reading these "computer interface" debates is a real revelation -- on how strategically neat was Linn to go the shortest way towards simplicity, robustness and sound perfection: ethernet streamer. NAS-in -> Sound-out. That's it.
I wish nDAC had this feature built-in.

Posted on: 30 July 2010 by james n

quote:
Reading these "computer interface" debates is a real revelation -- on how strategically neat was Linn to go the shortest way towards simplicity, robustness and sound perfection: ethernet streamer. NAS-in -> Sound-out. That's it.

But via a rather awful control interface when first introduced

I was looking at the unitiserve when i picked up the 252. If i was going the Naim DAC route then that would be my 'streamer' of choice - very neat.

Posted on: 30 July 2010 by pcstockton

quote:
Originally posted by AMA:

I wish nDAC had this feature built-in.

I bet the Linn owners wished there was at least one spdif input on their DS player.

Posted on: 30 July 2010 by pcstockton

quote:
Originally posted by AMA:
NAS-in -> Sound-out. That's it.

Ah you mean just like the HDX, NS01, Uniti, Uniserve and Unicute?

Is it such a big deal that Naim has an approach with a more expandable, hub style option?

Is it simply the box count that you have a problem with?

Posted on: 30 July 2010 by Hook

quote:
Originally posted by AMA:
Reading these "computer interface" debates is a real revelation -- on how strategically neat was Linn to go the shortest way towards simplicity, robustness and sound perfection: ethernet streamer. NAS-in -> Sound-out. That's it.
I wish nDAC had this feature built-in.

Hi AMA -

I have not heard anyone singing the praises of Linn's software. An integrated approach can be nice, but only if all the components of the solution are class-leading.

Am glad that Naim made the DAC a separate device, and gave me the freedom to choose my own front-end, and my own software interface.

Also, have not yet seen a reply about how any of these alternative interfaces would result in better sound than what we have now with the Naim DAC. Maybe there is a good argument, or maybe this is simply an academic discussion.

Hook

Posted on: 30 July 2010 by tonym

quote:
Originally posted by pcstockton:
quote:
Originally posted by tonym:
There's no doubt even Firewire is a far less satisfactory means of digital transmission than 12S but the latter won't work over a distance greater than a few cm.

A few CM? The i2s fitted DAC, i2s mods, and USB>i2s converters from the likes of Empirical Audio, Stello, and Northstar severely disagree with that.

But that introduces another code/decode. My statement is correct for pure 12S.

Posted on: 31 July 2010 by AMA

quote:
Originally posted by ghook2020:

Hi AMA -

I have not heard anyone singing the praises of Linn's software. An integrated approach can be nice, but only if all the components of the solution are class-leading.

Am glad that Naim made the DAC a separate device, and gave me the freedom to choose my own front-end, and my own software interface.

Also, have not yet seen a reply about how any of these alternative interfaces would result in better sound than what we have now with the Naim DAC. Maybe there is a good argument, or maybe this is simply an academic discussion.

Hook

Hook, the time is flying and Linn DS range is equipped with variety of decent SW which don't make the users crying anymore. Apart from native solutions one can use established music servers like Slim Server for example. I use KDS frequently in my friend's system (with Kinsky interface on PC) and also with iPad and it's absolutely no problem to control it. The main point for me (and many other audiophiles) is still SQ which is FANTSASTIC with KDS. They solved the transport issues by eliminating a jitter from the beginning of the pathway and offering users a complete solution. Viewing a rack with HDX/DC1/nDAC/XPS/PLs which is more expensive than small KDS box and the auditioning them in the same system is a very ... peculiar experience.

I'm still lacking a Naim signature in KDS sound - for example nDAC/XPS bass has better weight and punch. And this means a LOT for me. But on 16/44 tracks KDS outclasses nDAC/XPS in smoothness and vinyl-alike presentation.
It's easier to listen for KDS for prolonged time, less fatigue.
Possibly my transports are not good (Logitech Transporter, NAD M5 dig out, HiFace) -- but I have a good experience of auditioning various transports through various DACs and I have a feeling what kind of the sonic signature they incur and I don't think a better transport will give nDAC more smoothness.

I guess those who value CDS3 more than nDAC or those who accustomed to vinyl presentation will possibly understand my point. For me nDAC is a temporary stop-over on the way to the reference Naim DAC. And I think it's a good idea to make it ethernet-ready.

Posted on: 31 July 2010 by DHT

I hope Naim will just buy in an async solution, whether they choose Weiss, Wavelength or M2Tech's async solution.

Posted on: 31 July 2010 by Hook

quote:
Originally posted by AMA:
quote:
Originally posted by ghook2020:

Hi AMA -

I have not heard anyone singing the praises of Linn's software. An integrated approach can be nice, but only if all the components of the solution are class-leading.

Am glad that Naim made the DAC a separate device, and gave me the freedom to choose my own front-end, and my own software interface.

Also, have not yet seen a reply about how any of these alternative interfaces would result in better sound than what we have now with the Naim DAC. Maybe there is a good argument, or maybe this is simply an academic discussion.

Hook

Hook, the time is flying and Linn DS range is equipped with variety of decent SW which don't make the users crying anymore. Apart from native solutions one can use established music servers like Slim Server for example. I use KDS frequently in my friend's system (with Kinsky interface on PC) and also with iPad and it's absolutely no problem to control it. The main point for me (and many other audiophiles) is still SQ which is FANTSASTIC with KDS. They solved the transport issues by eliminating a jitter from the beginning of the pathway and offering users a complete solution. Viewing a rack with HDX/DC1/nDAC/XPS/PLs which is more expensive than small KDS box and the auditioning them in the same system is a very ... peculiar experience.

I'm still lacking a Naim signature in KDS sound - for example nDAC/XPS bass has better weight and punch. And this means a LOT for me. But on 16/44 tracks KDS outclasses nDAC/XPS in smoothness and vinyl-alike presentation.
It's easier to listen for KDS for prolonged time, less fatigue.
Possibly my transports are not good (Logitech Transporter, NAD M5 dig out, HiFace) -- but I have a good experience of auditioning various transports through various DACs and I have a feeling what kind of the sonic signature they incur and I don't think a better transport will give nDAC more smoothness.

I guess those who value CDS3 more than nDAC or those who accustomed to vinyl presentation will possibly understand my point. For me nDAC is a temporary stop-over on the way to the reference Naim DAC. And I think it's a good idea to make it ethernet-ready.

Hi AMA -

First of all, let me say that I read all of your posts carefully. You put a lot of effort into research and comparisons, and that benefits this entire community. Thank you for that!

A couple of things I read in your response jump out at me. It appears that in order to get the Naim/Linn comparison on somewhat equal footing, you front-end stuff the Naim DAC with an HDX. Take the HDX away and replace it with a $1000 streaming solution that delivers bit perfect, and we are now comparing a Linn solution that costs well over twice as much.

It sounds like you are hearing differences among front-ends for the Naim DAC. I do not. Or if I do, they are so bloody small that the subtleties are lost on me. I think we all recall the marathon debate about front-end sources for the DAC sounding different. Correct me if I am wrong, but the end result I recall was a general concurrence that the only way that DAC front-ends could sound different was from the environmental effects of EMI/RFI, and those could be mitigated by separation and/or Toslink. Perhaps you reached a different conclusion.

Anyway, thank you also for framing this connectivity discussion in terms of future product. Will be interesting to see what Naim does to elevate performance in a "reference DAC". No clue if they will enhance their current architecture or start over again. All I can say is that if they choose to make their reference DAC an integrated solution by adding ethernet and software, I hope that they will not lose the S/PDIF input. I believe that no matter how good Naim gets at software, that they will be continually playing catch-up with companies that are 100% focused on streamers and/or PC-based music player software. Perhaps Naim will build two products: an integrated DAC and a reference DAC separate. Given the current marketplace however, I have my doubts about the former. It seems that space if flooded with low cost solutions like Sonos and SB, and that Naim is trying to compete with the Qute. We'll see.

I am curious about your "fatigue" comment. You and I have very similar setups. For digital music, we differ only in our speaker choices. I can say with 100% certainty that I never experience fatigue in long listening sessions with my Harbeth C7s. Maybe the difference is that the C7s have a slight roll-off in the treble. Have seen reviews that say your D28's were designed with a slight treble lift. Just a guess.

I also find your KDS as "more analog" sounding comment interesting. I wonder if Linn did a nicer job of engineering their output stage? If a DAC is being delivered bit perfect, and the DAC chip itself is not broken, then where else but the output stage can differentiation be engineered? Just curious.

I have heard the KS only one time, and it was front-ending a wildly different setup (top-end ARC amps and Magnepan panels). Sounded great, but that was not a surprise at that price point! Am surprised to hear that in your comparisons that the Naim DAC is delivering better extension than the KDS. I, too, prefer the sound of vinyl, but mostly for Jazz and other acoustic music where midrange is everything. I almost always switch to digital for rock, prog, etc., where the added extension makes/breaks the sound. So I think we are in sync there.

Lastly, have never heard the CDS3, but it sounds like a wonderful player. Irrelevant for me though. I get all of the tactile feedback I want or need by shopping for/cleaning/playing vinyl. CDs are strictly a buy/rip/store thing for me. I could not ever see giving up the fast search and DJ capabilities that streaming gives me. Would be glad to someday evaluate a reference Naim DAC separate though!

Hook

Posted on: 31 July 2010 by Hook

quote:
Originally posted by DHT:
I hope Naim will just buy in an async solution, whether they choose Weiss, Wavelength or M2Tech's async solution.

Because.....?

DHT -

Have tried a couple of times to get you to elaborate, but you seem very hesitant to do so.

Not trying to be pushy -- just very interested in why you are so convinced that these alternative connectivity solutions would bring benefit to a future Naim product. Would really appreciate a reply beyond "it's better" or "it's fact".

Most people join forums to interact with and exchange ideas with other forum members. C'mon - give it a try. Not trying to be pushy, just trying to learn from your experience!

Thanks.

Hook

Posted on: 31 July 2010 by Eloise

Async USB / FireWire:

The basic advantage of an async connection (USB or FireWire) is that the controlling clock is kept as close as possible to the DAC (chip) avoiding jitter and this is then fed back to control the computer transfer. Async transfer (as defined by Gordon @ Wavelength at least) also has 2 independent fixed clocks for 48 and 44.1 based sample rates - the Weiss utilises PLL to derive the clocks from the clock used for FireWire. I'm unsure if it's due to philosophical or technical issues but the Ayre and Wavelength DACs are pure USB (no SPDIF inputs). Arcam's rDAC on the other hand utilises dCS's implementation of async USB and has managed to include SPDIF alongside.

The problem (that I see) with this is that in the Naim DACs schema, the output of the buffer circuit is already meant to be independent from the input (in terms of clocking). In other words, there is already an async detachment between the SPDIF input and the DAC stage. Naim have already stated that having USB connection within a DAC would cause more troubles (PSU / EMI) than it may solve.

Eloise

Posted on: 31 July 2010 by Hook

quote:
Originally posted by Eloise:
Async USB / FireWire:

The basic advantage of an async connection (USB or FireWire) is that the controlling clock is kept as close as possible to the DAC (chip) avoiding jitter and this is then fed back to control the computer transfer. Async transfer (as defined by Gordon @ Wavelength at least) also has 2 independent fixed clocks for 48 and 44.1 based sample rates - the Weiss utilises PLL to derive the clocks from the clock used for FireWire. I'm unsure if it's due to philosophical or technical issues but the Ayre and Wavelength DACs are pure USB (no SPDIF inputs). Arcam's rDAC on the other hand utilises dCS's implementation of async USB and has managed to include SPDIF alongside.

The problem (that I see) with this is that in the Naim DACs schema, the output of the buffer circuit is already meant to be independent from the input (in terms of clocking). In other words, there is already an async detachment between the SPDIF input and the DAC stage. Naim have already stated that having USB connection within a DAC would cause more troubles (PSU / EMI) than it may solve.

Eloise

Thanks Eloise. I always appreciate your clear and helpful responses.

So given the goal is jitter reduction, then job done, right? Naim has achieved zero-addded jitter through buffering/re-clocking.

So does this whole discussion come down to simple plug compatibility, and the elimination of USB-to-S/PDIF converters?

Early this year, I followed your recommendation of the RME PCI card, and have been thrilled by the results. Freedom to choose my software interface. Automatic support of all of the bit depths and sample rates (without having to change any settings, or restart any software), and bit perfect delivery over S/PDIF. Even if you had not helped me go in this direction, would probably still be doing ok with the M-Audio Transit, or maybe would have tried the HiFace.

Still, if this conversation is about nothing more than simplifying the connectivity, and has nothing to do with actually getting better sound, then as a Naim owner, I still do not understand what the big deal is all about. As you said, Naim claims that USB brings its own set of challanges, at least in the context of their chosen architecture.

Is the case being made here that Naim would find wider acceptance, and not lose anything sound-wise, if its next DAC had USB input? I could certainly understand that. But it would still not give me a reason to seek an upgrade. Given my current setup, my decision would be based on solely on an improvement in sound quality. Unless I am still missing something, it does not sound like async USB improves sound quality, but is simply an alternative method for potentially getting to zero-added jitter (with perhaps simplified, more universal USB connectivity -- and assuming that the PSU/EMI issues you mentioned are addressed).

Hook

Posted on: 31 July 2010 by js

None are perfect but given the right parameters a few can approach it. Whats nice is that there are extremely good ways of getting a result when done well and analog was never perfect either. Your PCI connection is the best you can do on a computer if noise isn't an issue.

Posted on: 31 July 2010 by AMA

quote:
It sounds like you are hearing differences among front-ends for the Naim DAC. I do not.

Hook, thanks for the kind remarks. Whether you hear a transport difference or not depends on where do you put a focus during audition and are you psychologically prepared to value a difference as "the difference". Try to listen bass-rich track through Hi-Face/WASAPI and some CD/DVD player and focus on bass articulation. During this post I'm listening Acoustic Alchemy, American/English, 2005 and the difference is very pronounced: the bass through HiFace/WASAPI has more weight and punch. It's not a shocking difference -- but this is one of the very elements which normally make a difference between 5 K$ good CDP and 25 K$ very good CDP.
How do you value this in VFM terms is very personal.

All I can say is that small demo advantages of better CDPs show up a BIG break-through during the prolonged listening experience. They normally provide better involvement, presence factor and less fatigue. The last week I was doing a routine job and I was listening for LP12 during 8 hours continuously and I was not tired at all! My nDAC/XPS is on the next shelf -- but it didn't seduce me to change despite of it's obviously winning control comfort. That's what I call a "fatigue" factor.

Posted on: 31 July 2010 by AMA

quote:
I also find your KDS as "more analog" sounding comment interesting. I wonder if Linn did a nicer job of engineering their output stage? If a DAC is being delivered bit perfect, and the DAC chip itself is not broken, then where else but the output stage can differentiation be engineered? Just curious.

Hook, I have a strong believe that digital harshness emanates from D/A chip design, I2V and analogue filtering. The output gain stage has nothing to do with it and I believe it's the strongest element of Naims CDPs and DACs. Although designed differently (and mostly driven differently) from Linn it features classic Naim virtues as opposite to Linn house sound. This goes down to individual preferences and I'm definitely in the Naim camp here.

The I2V/filtering stages create MUSIC -- they define the primary analogue output from the D/A chip. Theoretically the correct I2V transformation and proper analogue filtering on 16/44 should reproduce the original waveform in 20 kHz bandwidth. Exactly the same original analogue waveform as it was digitized during ADC. Mathematically it's a proven fact. Of course the actual implementation of these electrical elements is far from mathematical formulas and this is the weak point -- and makes up to 90% of DAC R&D and commercial cost.

The hi-res domain benefits from less complicated filters and you can get high quality output much easier than for 16/44 bitstream. That's why (I believe) hi-res sounds so closer to vinyl than 16/44. That's why upsampling DACs often sound so much better than non-upsampling (although many simplistically upsampling DACs eventually spoil the other elements of music reproduction).

BTW I did not yet compare nDAC vs KDS on hi-res (plan for October) but nDAC is definitely smoother on hi-res than on 16/44.

I would also like to see if someone can test Paganini or Scarlatti upsampler with nDAC (they can work as USB transport as well) -- and I expect nDAC will produce a smoother sound working constantly in hi-res mode.

Posted on: 31 July 2010 by JYOW

quote:
Originally posted by js:
None are perfect but given the right parameters a few can approach it. Whats nice is that there are extremely good ways of getting a result when done well and analog was never perfect either. Your PCI connection is the best you can do on a computer if noise isn't an issue.

Agree with that, as most of us agreed, there is no hard and fast rules, the devil is in the details. Both Naim and Berkeley decided not to implement in-built computer interface for noise reasons, and it apparently does not bother Weiss and other who implemented their solutions with good results. Naim decided to go with S/PDIF buffering, which, apparently they are not the first in doing so, both my DAC2 and DAC202 do a bot of that, and I believe CHord has been doing that in a big way.

They all work very well in their own way, the rest is down to preferences.

Posted on: 31 July 2010 by js

As always.

Far from the first but a bit unique in topology and level of implimentation.

Posted on: 01 August 2010 by AMA

quote:
both my DAC2 and DAC202 do a bot of that, and I believe CHord has been doing that in a big way.

JYOW, do you mean that DAC2 and DAC202 re-clock input bitstream through saving to temporary memory buffer?

Posted on: 01 August 2010 by JYOW

quote:
Originally posted by AMA:
quote:
both my DAC2 and DAC202 do a bot of that, and I believe CHord has been doing that in a big way.

JYOW, do you mean that DAC2 and DAC202 re-clock input bitstream through saving to temporary memory buffer?

Yes I am quite sure they do. I had problem with lipsync delays on my DAC2 with SPDIF and when talking to Weiss they mentioned it was due to the fact that the data is pre-buffered.

In fact, in the previous firmware I believe ther was an option to adjust buffer size in the Weiss control panel. If you Google Weiss DAC2 and Buffer you will see an article from Asia Weiss (Kent Poon) writing about DAC2 with a screen shot of the buffer size option.

Posted on: 01 August 2010 by Eloise

quote:
Originally posted by JYOW:
Yes I am quite sure [the DAC2 and DAC202] do [have a buffer]. I had problem with lipsync delays on my DAC2 with SPDIF and when talking to Weiss they mentioned it was due to the fact that the data is pre-buffered.

In fact, in the previous firmware I believe there was an option to adjust buffer size in the Weiss control panel. If you Google Weiss DAC2 and Buffer you will see an article from Asia Weiss (Kent Poon) writing about DAC2 with a screen shot of the buffer size option.

Is the adjustment of the buffer size in the Weiss control panel not related to computer side buffer rather than buffer in the DAC?

Eloise

Posted on: 01 August 2010 by james n

quote:
Is the adjustment of the buffer size in the Weiss control panel not related to computer side buffer rather than buffer in the DAC?

Yep - Computer not DAC.

James