A comparison between two DACs, one R-2R, the other ES 9038-based.


This is an item from the FWIW Department, I guess.

Recently I set up an A/B comparison between a Denafrips Pontus and an OPPO Sonica…. Both DACs fed from a Nuprime CDT-8, via the XLR inputs on the same preamp, and everything else constant through the two signal paths to the speakers.

The result when I repeatedly toggled back and forth from the preamp’s input one [Denafrips] to its input two [OPPO]?---- To my surprise [and disappointment], the sounds from the two DACs were utterly indistinguishable, across all kinds of music, after repeated trials … impossible to tell apart… impossible.

The moral of the story? I don’t really know, but it does suggest to me that those who say that DACs of comparable quality cannot be told apart just might have a point.

I bought both the Pontus and the Sonica because I thought that it would be nice to have on hand DACs of “different flavours,” one based on an R-2R ladder, the other based on a delta-sigma chip. 

I did want the expected difference to be real… just for the fun of it… else why spend the extra money? So, my “confirmation bias” was, if anything, stacked in favour of there being a detectable difference.  

However, the results of a reasonably well controlled comparison [sadly?] did not bear out that expectation. Differently based DACs, 2-R2 vs delta-sigma, may not offer such different flavours as many suggest. Is that claim all much ado about nothing?

Thoughts from members of the Forum?

 

 

 


128x128Ag insider logo xs@2xzimmerma

Showing 1 response by cleeds

wtf
... I've found the A/B quick-switch method doesn't reveal DAC nuances nearly as well as a lengthier listen. For me, a relaxed non-focused listen allows me to hear the subtle differences and, once heard, I can never un-hear them ...
I agree and think it's true of not just DACs, but audio components in general. That is why I find the concept of an audio "shoot-out" (as some promote here) to be so useless. That's not to say that quick A-B switching has no value at all, but it isn't the best way to evaluate an component by a long-shot, imo.