Measurements and Review of Schiit Yggdrasil DAC

[Jimster480]

They do have pretty low noise for an R2R DAC.

Compared to what other modern R2R DAC? Because if you compare it to a soekris DAC.... well Soekris DAC's are better measuring.

So I mean they are claiming to measure better, and be overall superior but we just haven't found that to be the case.

 

[mindbomb]

Perhaps they thought it sounded good? The relatively poor measurements, which I don't really think they could be considered poor, were either inaudible or did not impact the sound negatively?

I think that is probably how they rationalized releasing this product. But if you go back to when they were conceptualizing the yggdrasil, it sure looks like they had decided to use that dac chip primarily due to marketing potential and not audio quality.

 

[garbulky]

I think that is probably how they rationalized releasing this product. But if you go back to when they were conceptualizing the yggdrasil, it sure looks like they had decided to use that dac chip primarily due to marketing potential and not audio quality.

I mean if you want to assign some sort of negative motive, that's always a possibility...or maybe they spent five years developing the best sounding dac they knew how to do as they've repeatedly said.

 

[Jimster480]

I mean if you want to assign some sort of negative motive, that's always a possibility...or maybe they spent five years developing the best sounding dac they knew how to do as they've repeatedly said.

Except that its not lmao

 

[mindbomb]

I mean if you want to assign some sort of negative motive, that's always a possibility...or maybe they spent five years developing the best sounding dac they knew how to do as they've repeatedly said.

Whether you look at the results from atomicbob, headfi, ASR, or stereophile, you aren't really seeing great performance from the yggdrasil. And when you note that it uses an unique multibit dac chip, I think the natural conclusion is that they compromised on performance in order to incorporate it, so they would have a dac that would stand out more.

 

[Jimster480]

Whether you look at the results from atomicbob, headfi, ASR, or stereophile, you aren't really seeing great performance from the yggdrasil. And when you note that it uses an unique multibit dac chip, I think the natural conclusion is that they compromised on performance in order to incorporate it, so they would have a dac that would stand out more.

Bingo.

 

[Frank Dernie]

Perhaps they thought it sounded good? The relatively poor measurements, which I don't really think they could be considered poor, were either inaudible or did not impact the sound negatively?

There are several reviewers in the media who always prefer equipment with high levels of distortion and/or frequency response altering levels of output impedance. It begs the old question do they like it because of the poor measurements or despite them? For me the only plausible explanation is "because" since if these effects were inaudible the item would sound no different would it? Also, if an item is very expensive and sounds a bit different it is human nature to assume the different is "better" just because it is so expensively achieved.

 

[Jinjuku]

True but we all have our moments , I just hope the guy saves a thought for the totally unpaid

Some seem to have a lot of 'moments'

 

[garbulky]

There are several reviewers in the media who always prefer equipment with high levels of distortion and/or frequency response altering levels of output impedance. It begs the old question do they like it because of the poor measurements or despite them?

I think it's that they like how it sounds? Most of these reviewers don't measure equipment or use it as a significant yardstick.

For me the only plausible explanation is "because" since if these effects were inaudible the item would sound no different would it?

Well that's a whole different question that probably should go in to another thread. But since it's a question, I would venture to say that these measurements have little to do with exactly how it's going to sound to someone - unless they are off by a good bit. I hope that in the future there will be more developments happening in correlating measurements to a more specific impression of the sound perceived.

Also, if an item is very expensive and sounds a bit different it is human nature to assume the different is "better" just because it is so expensively achieved.

Here's how I think this happens. The price difference doesn't matter a whole lot to them. So when there is a $15,000 piece of gear and a $500 piece of gear and they sound similar or only slightly different. You'll notice there's little critique given to the extreme price difference. Probably because they are not the ones paying for it. So if they hear a difference, that's "good enough".
High prices and average performance are rarely penalized in reviews, I think mainly for this reason .There's no personal stake in taking a manufacturer to task. They see tons of gear go through their office, this is just another one of them. They didn't save up for six months to buy it.

 

[Frank Dernie]

Well that's a whole different question that probably should go in to another thread. But since it's a question, I would venture to say that these measurements have little to do with exactly how it's going to sound to someone - unless they are off by a good bit. I hope that in the future there will be more developments happening in correlating measurements to a more specific impression of the sound perceived.

I am firmly of the opinion that something sounds different because of the measured deviations. The idea that these distortions/additions are actually inaudible but the device sounds better for some magic as-yet-to-be-discovered reason is totally cockeyed IMO.
Audio is a mature science and there is not much genuinely new being brought out, either in the equipment or the understanding of the engineering. DSP for speakers and rooms is about it.
I was involved in R&D into the design of record players in the 1970s. I have seen nothing around today about record players that wasn't fully understood back then, and a lot of total BS/misunderstanding from enthusiasts which would exasperate those who were active back then since so much has been forgotten or not learned by a lot of present day practitioners.

 

[garbulky]

I am firmly of the opinion that something sounds different because of the measured deviations. The idea that these distortions/additions are actually inaudible but the device sounds better for some magic as-yet-to-be-discovered reason is totally cockeyed IMO.
Audio is a mature science and there is not much genuinely new being brought out, either in the equipment or the understanding of the engineering. DSP for speakers and rooms is about it.
I was involved in R&D into the design of record players in the 1970s. I have seen nothing around today about record players that wasn't fully understood back then, and a lot of total BS/misunderstanding from enthusiasts which would exasperate those who were active back then since so much has been forgotten or not learned by a lot of present day practitioners.

I'm going to take this to another thread so as not to clutter this thread.

 

[L0rdGwyn]

AtomicBob has posted some new measurements of the Yggdrasil version 2 on HF:

https://www.head-fi.org/threads/schiit-yggdrasil-v2-upgrade-technical-measurements.881867/

Looks like they cleaned up some Schiit up from version 1 to version 2, the linearity and 1 kHz sine look to be improved.

 

[mindbomb]

iirc, the linearity isn't better, they just zoomed out to make it look smoother. I will say that the 121-122 db on the -60db dynamic range test on balanced (which was in previous atomicbob measurements, but I don't think was brought up enough) is a legitimately good figure, just to give credit where credit is due.

But the smpte imd results were done at a suspiciously low voltage, and if you look at the bryston bda2 review here, I think you can see why. The full scale results are probably pretty ugly. And the ccif imd results were literally cut-off on the x axis, also conveniently before full scale. This strange imd performance is in the old stereophile review too, so it's apparently been there a while.

 

[L0rdGwyn]

Hmm, what I am seeing is both V1 and V2 on the same plot, with V2 nearly hitting -120 dBFS while V1 is identical to AB's first measurements, coming in around -95 dBFS, consistent with Amir's original measurements as well. Maybe I am misinterpreting.

 

[mindbomb]

Well, the -90db sine wave looks actually really good on his v2. I can't tell about the linearity with the scaling how it is.

I'm rereading Jude's post about linearity, and I don't get something. He said amir didn't measure the balanced out and thus got bad measurements. But on unbalanced, Jude got -95db, within 1db, while amir got -98db within 0.1db. So Amir's results were substantially more linear. I took for granted Jude's claim that the ASR linearity measurements were unbalanced... but were they?

I'm feeling like maybe these questions about linearity discrepancies were answered already in amir's response to jude, but I don't want to dig through a schiit thread on headfi lol.

 

[amirm]

AtomicBob has posted some new measurements of the Yggdrasil version 2 on HF:

https://www.head-fi.org/threads/schiit-yggdrasil-v2-upgrade-technical-measurements.881867/

Looks like they cleaned up some Schiit up from version 1 to version 2, the linearity and 1 kHz sine look to be improved.

I wrote a detailed response and alas, they are still "moderating" my posts and won't let it go out as any other person's post. So much for wanting to have a proper discussion about their results. Here is what I wrote, just in case they edit or delete it:

----

While his measurements might be explained by a bum Yggdrasil 2 unit

Ah, no. Hell no.

Before I get into that, let me say that it is wonderful to see more objective measurements and data in this forum. So for that, I am thankful of your efforts and that of Bob Smith's OP.

That said, I wish you would just focus on the measurements, read what I had post originally to which your responses are about, and then we could discuss them. Here is my original measurement and review: https://audiosciencereview.com/foru...ents-and-review-of-schiit-yggdrasil-dac.2358/

In there, you see that my access to the *two* (yes, two (2)), Schiit Yggdrasils came about because our local audiophiles and friends had them and kindly allowed me to measure them. Here is the picture of the basement of the church were we have our meetings showing all the equipment:

If you look in the rack on the left, uo see the Schiit Yggdrasil which had a Gen 1 audio board (but the latest USB board upgrade).

If you look to the top right of the rack, you see the second Schiit Yggdrasil which had the Gen 2 audio board (but first gen USB).

Also visible is the Berkeley alpha DAC and later on (not in the picture) the Bryston BDA-2 DAC. Again, all of these were brought my friends at this monthly gathering of our audiophile group.

So I not only measured two Yggdreasil DACs, but also two others in identical setting and in front of 20+ people. Last thing I wanted to do was to tell my friends there that their box did not perform. So any bias I had here was to tell good news to everyone who let me measure their gear. Alas, it did not work out that way. I finished my measurements and made a quick powerpoint which I presented to the group. These are the slides from that which match the review and measurement which was the result of a yet another trip to re-confirm the Schiit measurement.

In all cases I had my measurement project file ready, move the cables from one DAC to another and press measure again. Any errors in instrumentations would directly reflect in all the DACs. Here is the residual noise and distortion of a 1 kHz tone where the tone itself has been filtered. Using a special analysis mode, almost all of the noise and distortions of the analog to digital converter inside the Audio Precision analyzer are eliminated, leaving us just what the DAC produces:

As you can see, it is abundantly clear, pun intended, which is the better performing DAC. The Schiit Yggdrasil has heaps of harmonic distortion on top of a ton of random distortion spikes. This is why its THD response is so much worse. In other words, the two measurements agree with each other.

This is how you confirm what you are seeing to be correct. You cross correlate between measurements of the same product and different ones. The Schiit Yggdrasil clearly has low level nonlinearities and errors that result in such spray of distortions.

Next slide was the noise and jitter products, again comparing the Berkeley to Yggdrasil DAC:

Looking at the Schiit Yggdrasil response around the main tone at 12 kHz, we see those red jitter spikes. They are symmetrical so are either clock jitter or reference voltage modulation. Again, both units were measured in the same sitting. Only the cables were moved and the identical test, with identical settings was re-run.

Next slide was linearity:

You keep complaining about my 0.1 dB threshold of error. Forget that for a moment. Just look at the overall graphs. By the time we get to the end at -120 dB, the Yggdrasil has gone of the rails big time. The Berkeley while demonstrating errors, is in much better shape than Yggdrasil.

In my review, I showed that this was true of both Schiit Yggrasil DACs:

We see the same wild loss of linearity at around the same point. So no, this not a murphy's law of one unit being bad. Both units with completely different pedigree, both of which had been through Schiit's factory twice with double the opportunity for their performance to be confirmed, producing non-competitive linearity, distortion and jitter results for this price class or frankly any price class.

Here is how the Bryston BDA did:

What we see on the right with Byrston, is what any DAC that is half competent and priced as low as $200 should do. Not what we see on the left from Yggdrasil DAC.

I will repeat one more time: all of these measurements were made in the same sitting, in front of the owners of said gear, and with identical measurement setup. Hell, you can tell from my misspelling of AudioScienceReview.com above that I used the same templates for all.

Based on this data, this was my concluding slide to the audiophile user group:

I was very concerned when I was typing the above slides that I would so offend the owners of the Schiit Yggdrasil that they would never want to talk to me. Credit to both, not only did they stay friends, they allowed me to go and measure their Yggdrasil DACs a second time to confirm these results. And confirm I did.

And oh, sitting in the audience was the chair of the user group and one of the designers of the Berkeley alpha DAC! As you see, despite my friendship with him and the rest of Berkeley crew, I had no choice but to declare the Berkeley not competitive either.

In summary, I know the measurements of Schiit products are always controversial since many of their owners don't want to know it has any faults. So I do everything I can to verify what I am about to say about them is correct. And I feel that I did that here. And your guess that start this missive from me , simply doesn't hold water. Two DACs with two discrete channels of audio all being randomly bad? No.

In the next post I will address your measurements and differences.

 

[palamudin]

Pro tip. Buy Sabaj Da3 :wink: screw these mega machines. Buy some quality wine for the change.

 

[Jinjuku]

This is why you run your own site folks. The best content will generally show up in search engines. Reasonable people will make up their own minds and the 5 unreasonable people will just stay at Head-Fi.

 

[rebbiputzmaker]

This is why you run your own site folks.

Yes, it is best to keep all this great knowledge confined to the few that can properly understand it. Leave the clueless to enjoy their music with their faulty equipment.

 

[rebbiputzmaker]

I was very concerned when I was typing the above slides that I would so offend the owners of the Schiit Yggdrasil that they would never want to talk to me. Credit to both, not only did they stay friends, they allowed me to go and measure their Yggdrasil DACs a second time to confirm these results. And confirm I did.

But you do not say what they think of your results, and if any of this changed their minds about the sound of their systems. Since you do not mention this at all, I might guess more than likely not. Since you probably shower regularly and use a good deodorant, why would they be offended, and never talk to you again?