Schiit Yggdrasil Stereophile Review + Measurements

As I recall (and I could be talking crap, never rely on my memory), the Yggdrasil uses discrete multibit DAC chips.

(Sneaky googling)

Yes, it uses four of something like these (two per channel):

http://www.analog.com/media/en/technical-documentation/data-sheets/AD5791.pdf

...to do the hard work. There's no external resistor ladder, I think. Very nice, accurate DACs, a bitch to use for audio (unless what little I understand of the above has mislead me). Smartarses.

(The Gungnir Multibit uses a very similar part, slightly lower-end.)

 

In that first diagram, I think that there is an r2r ladder inside the block marked "20-bit DAC". If I'm understanding it correctly, it talks about it on page 19.

 

Yeah, looks like it's a hybrid, with the fourteen LSBs being an R2R ladder, and the six MSBs being discrete resistor (strings?). Shiny.

Edit: Ninjaed by Marvey, yay. Thanks!

So @Marvey, is that "pulse array" stuff in the Chord DACs effectively like Delta/Sigma on some level?

 

Yep, this is why I'm asking you, not John Franks (also, you can form a coherent sentence). There's a lot of hand-waving vagueness which gets enthusiastically parroted.

The FPGA seems to mostly be doing duty for what would be a DSP in most DACs- I guess that's a relatively power/heat and labour-intensive way to go about it? Chord seems to like to roll their own subsystems quite often just for the hell of it- though maybe it's harder to run eleventy-billion taps on most sane DSPs?

Sorry, I should stop threadjacking, but curiosity got the better of me.

 

How worthless does this make the Stereophile archives for comparison purposes between older and newer gear?

I, for one, am saddened by this turn of events.

 

I don't think any tech magazine uses the same measurement testing now that they used 15 years ago.

 

[Sorry ignorant noob post removed]

 

I recommend you introduce yourself here:
http://www.superbestaudiofriends.org/index.php?threads/new-members-introduce-yourself.17/
Read the first post there too.

 

[Sorry ignorant noob post removed]

 

Yes! They were very tight about class A. Back then I bought some used tubed equipment based on their class A rating. Still own them. Been with me through many speakers. As time has gone on, so much is class A that doesn't come close to the quality or sound of my old equipment that I have lost faith in objective reviews from them.

 

I don't think I ever saw anyone recommend a Hugo 2 over a Gungnir Multibit let alone the Yggdrasil.
It's priced as much as the Yggdrasil, yet has no AES, XLR etc...

I always got the impression that it competed with the $500-ish DACs & not the $1k & up stuff.

 

Did someone miss the sarcasm??

 

Inside joke.

 

Even the waveform he achieved by cheating seemed alot more jagged all over compared to the Yggdrasil results they posted.
I can't imagine what Dave would have looked like had he used same method settings for Yggdrasil.
I will try to ask.
Edit :
Ok, I asked. . Tried to be "diplomatic" about it, but,

Doubt I get an honest response if at all.

 

If they are, they're being silly - since outside of some pre-production units, with non-final programming, I don't think anyone has actually has heard the Hugo 2 outside of a couple of recent shows (which is, of course, no place to be making critical evaluations).

What surprises me is that, in the last four or five days, between here and Head-Fi; I've gotten about a dozen PMs asking about whether I'd buy Hugo 2 or Yggdrasil, or what I think of Hugo 2 in comparison to other DACs. How about we wait till the damn thing ships - given that it's now at least two months late, and June isn't looking good either?!

But while on the subject (even if somewhat off-topic) ...

Given that I found DAVE to just barely edge out Yggdrasil, and the margins for that have gotten slimmer every time I've spent more time with DAVE, it's hard to see taking anything away from that performance and still having the unit be competitive. With Hugo 2 having half the number of pulse-array elements, and running much shorter tap-lengths*, while I'm sure it'll usefully clobber both Mojo and Hugo, it's hard to see it measuring up to Yggdrasil.

As ever, the proof is in the listening ... but anyone recommending Hugo 2 today is almost certainly doing so on the basis of never having heard the final product.

--

*The big tap-length numbers are amusing. Greatly simplifying things ... you can only only make the filters so big before you exhaust your conversion coefficients. When you over-sample to 256 Fs and then to 2048 Fs, as the Chord DACs do, this extends the number of taps you can (and should) apply. So while it's a prominent feature of their marketing, it's not like other manufacturer's are necessarily being lazy here ... it just won't work without the extreme levels of oversampling at work here.