DENAFRIPS Terminator and Schiit Yggdrasil Evaluation and Comparison

@david_ten, thanks for a very comprehensive review and I’m not surprised that it betters the Yggdrasil. Other forums posters have compared the Denafrips favourably with more expensive dacs such as from Total and dCS. If you get a chance I suggest you try a Mola Mola dac, the MM would be a step up again in performance I believe.

@jayctoy Thank you. Do you know if Denafrips will be actively shown and demoed at Axpona?

@toteapaudio Thank you for you comments. I will make an effort to listen to the Mola Mola here in the US. Where are you located in the U.K.?

BTW, I love the simple and elegant layout of your site. Beautifully done. Boenicke speakers are already on my list. : )

I don’t doubt the Denafrips sounds fantastic. However technically I have a bit of trouble with the outlandish claim of 26 bits resolution. I smell a rat that may have been released by the Denafrips marketing department. (Not unlike Yggdrasil which have dubious technical claims also - usual audiophile marketing hyperbole as some would regard this)

26 bit capability requires precision of 1 in 67 million. This resolution requires a tolerance of 0.0000015%on the resistors.

The designers claim they actually use 0.005% tolerance resistors which is technically between 14 and 15 bits resolution.

The only thing that could solve this fact of physics or mathematics is if the designer is using a very large number of resistors and randomizes the selection every clock cycle. In theory you get a reduction in noise by the square root of the number of resistors.

Applying mathematics again, to get to 0.0000015% accuracy (26 bit resolution) from 0.005% tolerance on an individual resistor would require 11 million resistors.

Can anyone with more technical knowledge about the Denafrips Terminator design confirm if they are randomizing more than 11 million resistors to create an output per individual channel? Overall the design should have at least 22 million R-2R resistors for two channels in order to meet the stated resolution of 26 bits - can anyone confirm this is what they have done?

Now that Mr Spock has fully weighed in, I wish to thank you for this review.

As with all things audio, the proof is in the listening and you have captured this in your writing. Congrats!

Dave