DAC Shootout Starts This Weekend

Okay...the Wyre4Sound DAC 2V2 SE arrived on trade.  After a little bit of time playing and getting settled, I ran the same test but first a few notes. This DAC has an insane number of settings. 

Unfortunately, they are in a menu that is accessible when the DACs main power is on but the DAC is not running.  This makes making changes more difficult but you have an immense amount of flexibility to fine tune your sound.  

This could take hours and bluntly, I don't have those hours to spend so I made a few tweaks to get the DAC into the system and the thing I found most intriguing was the jitter eliminator though I don't love the net result.  It sounds less like true, proper clocking and more like a DSP correcting for a clocking error which it probably is. 

Soundstage is great.  This blows the Hugo 2 and AVM out of the water in terms of imaging and staging.  Overall, detail was very good.  I puled out the Audiobyte and Hugo 2 as points of comparison.  

Detail in drum rolls are strong but a hair below the DACs in the price class above.  Well below the TT2 or the Audiobyte.  There is a touch of sibilance in Sting's voice and in certain parts of Liberty.  Two or three moments of glare during Liberty that were tough to listen too.  This is likely based on the setting I quickly arrived at.  

Strings were slightly massed in the Berg piece and due to the size of the soundstage, the intimacy was lost.  The opening of Duende was a little bit of a struggle.  It did the job, but not at an elite level. It sounded a little muddy.  

It is priced in the middle between the Audiobyte and the Hugo 2.  Performance is in the middle between the Audiobyte and the Hugo 2. Detail is closer to the Hugo 2 while staging is closer to the Audiobyte.  

At $3800, as a new DAC, it is priced fairly and delivers very good performance if you are wiling to spend the time tweaking in what is not the best user interface.  Used, in the $2k range...this thing is incredible.  

it is much more the implementation of the individual design than the basic architecture (R2R, FPGA, Delta Sigma)that differentiates dacs: i.e. quality of clocking and power supplies( foremost), quality of the noise rejection on digital inputs, design of attenuator (if used as a preamp), presence of op-amps or capacitors in the signal path, direct coupling of output stages and on and on. I very much question the ability to differentiate between the different chip sets, be they TI/Burr Brown, AKM, ESS Sabre et al. in an optimised setting on sound quality. Assuming optimal clock accuracy it is really the differences in the analogue domain that determine dac sound quality.

agree with @antigrunge2 100%

I agree regarding implementation. There are some limitations with R2Rs and DSD. Otherwise, the only thing I can see against any design is if a DAC uses a particular chip that is discontinued and/or replaced, the older models will drop in perceived value. That obviously doesn’t happen with R2R DACs or FPGAs.

Couple of updates.  

I have done some testing with cables.  The AQ Diamond is very good.  I swapped in a Clarus Crimson and think it was a pleasant switch depending on the speakers and is a very good substitute.  if you have issues with box on AQ stuff fitting, this is a brilliant substitute.

In some cases, I reported a bit of bass emphasis with some DACs.  Swapping speaker cables, I think that was the AQ Robin Hoods, not the DACs.  It is definitely a more bass heavy cable which is probably why I like it with my stand mounts.

Finally, I have done some additional testing with different servers.  I have an Antipodes K40 here (and an open box) along with my sample Nucleus+, a traded in Nucleus and a Wavedream NET that was just traded in with a Wavedream Signature DAC.    

The Nucleus+ and Nucleus exhibited no difference as standalone servers.  I do think there was a small improvement with the Nucleus+ when they were connected via USB and the Roon unit was used as server and renderer but it was small enough that I am not certain and it could be my internal bias.  

The Antipodes K40 as a server used either with the Weiss directly via ethernet or with the Chord (2Go/2Yu) renderer is significantly better.  It should be at 4 to 6x the cost of the nucleus but everything is clearer and cleaner.  Effects, especially in Liberty, clustered notes like the opening of Duende or bells in Be Still My Beating Heart are just clearer. 

The Wavedream NET as a Server vs the Antipodes were a push when feeding the 2Go/2Yu.  Feeding the Wavedream Signature takes this DAC to a totally different level.  The loss of that last, tiny bit of jitter is just mind boggling.  I am using an Audioquest Dragon 48 HDMI to connect them.  The idea is that the clock is perfectly synchronized between the two devices and that is completely obvious.  

 

So is there a 1-5 ranking coming out of this?