Schiit Yggdrasil Stereophile Review + Measurements

Truncated? No, rounded after DSP math at much higher precision.

The only thing truncated are our prices.

Sorry, couldn't resist.

 

f**k it, life's too short to worry about the biases of audiophile press. I'll just reenable my adblocker on Stereophile to show my displeasure. Don't worry @Tyll Hertsens , you'll still get the ad revenue on Inner Fidelity when I visit, and MF will get the same on Analog Planet.

Also, I hope you see a spike in sales when the new add rolls out @schiit

 

On the other hand, I have to thank John Atkinson for giving us a great theme for an upcoming ad series, starting with:



Yep, we're obsolete.

We use tubes.
We use multibit DACs.
We use discrete design.
We use Class A and AB topologies.
We use mechanical switches and relays, rather than touch screens and apps.
We even use outdated manufacturing techniques, like sheet metal fab.
And we're crazy enough to insist on doing this in the USA (long before it was fashionable.)
Aaaaannnddd...our prices are straight out of the 1970s!

Yes, we're obsolete. And we're 100% fine with that.

 

Given that one of the state of the art Dan D'Agostino amps with it's milled copper heat sinks had catastrophic thermal failure when JA was testing it, I think you guys have the right idea.

 

I love the parenthetical addition at the end of the obsolete ad. Keep disrupting the game!

 

Man, what a mess this guy is making for himself on social media.

 

Not quite. Truncation would be just shortening the word depth. Rounding or saturation or other more clever techniques would be different process.

 

That was my point.

 

My misread.

 

I feel like there's some witty joke to be made about competitors and truncation, but that it would be wasted on anything except one of @baldr 's trade show shirts.

I need to go look if JA does any testing with DSD files, and see what he has to say about 1-bit conversion. I bet there will be some grade-A hypocrisy.

 

Does INL/DNL affect human tonal perception of DACs too and not just ENOB? Stuff that's obviously warmer sounding than neutral in the upper bass/lower mids measures flat on linearity tests yet discerning listeners can perceive the almost-oversatured sounding inner-warmth in stuff like ESS Sabres, most AD1955 implementations, and the Modi Multibit/Bifrost Multibit rather easily.

 

Should also include an elephant (in the room) and something about trunk-ation [sic] .....

Ps. I'll find myself out. DS.

 

The stereophile article is by Herb Reichert, not JA. Really have no idea if he's in charge of the Facebook messages.

Edit: found measurements page. Forget I said this

 

Marvey, looking at your data and the data from JA I do see the same pattern emerging. Their is some mixing of the 19,20khz signal making spurs along side the main tones. I am not saying that it is bad, it looks very well controled. Compared to the other DAC mesuements shown it is far more linner, showing that the filtering and the analog is of the highest quality. These spurs are more a product of transistor non-linearity then of how long the digital word converted was.

Edit: sorry if I miss understood the point of their beliefs that the noise floor they were measuring was periodic noise and not random noise.

 

A good thing the Schiit heads have addressed these points well before Stereophile's take. @baldr on truncation/rounding:

and more:

So Stereophile's "sound of truncation" makes me wonder if they know what they're talking about, or simply talking out of their collective arses...


And now onto 24-bit data... Again, @baldr:

Just one of his rants on the subject. My understanding is that with PCM you do not get linearity above 2^20, or just over 1 million levels (1048576 to be exact). As such, ALL so-called 24-bit recordings have some 4 effectively unused bits.

Or, in the words of the man himself:

So once more, I'm wondering if John Atkinson knows what he's talking about...