Music Server/DAC – Ethernet, USB, S/PDIF, AES3 or I2S.

Ignores the fact that some DAC designs are almost immune to recloking.

Antipodes makes good argument for their K50 solution. Ed Meitner of EMM Labs has stated having clock closest to circuit it's clocking is preferable, in this case we'd want to use the clock in dac.

 

I'd love to to a direct comparison.

@sns

I have long admired and valued Ed’s contributions. That’s one argument to have clock close to its circuit for best performance but I can’t stress enough the improvements in precision and realism one hears from a SOTA external clock, assuming your downstream components are good enough. In a direct comparison between DA2 and Merging +NADAC and +clock (external clock), I was in a different stratosphere. As good as DA2 is on its own, it simply paled in comparison. This revelation came after enjoying DA2 with Aurender N20 for over two years.

My take away is, unless one have the opportunity to experience and compare, we are speculating at best. In audio, the best in our minds is what we can afford at present. I am only using a external clock that cost $18.5K. The next two level of upgrades in Merging clock are $30K and $58K :-)

@lalitk 

“ … In audio, the best in our minds is what we can afford at present.…”

+1 

 

 

 

@Akg_ca I prescribe to the PS Audio architecture with a Bridge II network card In my DAC hitting a NUC rock fed by Tidal and Qubuz. I have electrically isolated the DAC network card by converting to optical then back to LaN right before the last 1ft of LaN. The optical to wire box is powered by ifi pure. I believe Paul's theory about digital feed and am wary of reclockers upstream of a high quality DAC which will have it's own clock. I believe the I2S is the best protocol and understand it works best as a direct connection i.e. no cables. Probably get some pushback from some who have spent big on cables and reclockers...