Measurements and Review of Schiit Yggdrasil DAC

[amirm]

AP helped set the test up according to Jude.

Not quite. I explained how my measurements were made to AP folks after pointing out to Jude that he was doing it wrong. They then worked with Jude on an implementation even though I had told Jude that the right solution on APx555 requires much more research. I have explained all of this to AP folks this morning again and issues in what Jude has published.

In a nutshell, they have set up a test where all the distortions and aberrations of the DAC are wiped clean. And they then declare: "oh look, it is linear down to -130 db" or whatever. Well duh. Of course if you remove all the noise and distortion from a DAC, it then looks accurate. Why bother running such a test when in real usage of the DAC no such filtering exists. Talk about running off with measurements with no thought of what they really mean and what benefit and correlation there is with audibility.

On top of that Jude continues to block all of my posts on head-fi. So no way for me to convey this information there. This is not the way we converge to a consensus.

So my advice remains: please wait to draw any conclusions until I remeasure the device with my APx555. Until then, my data remains 100% valid in pointing out serious issues in the performance of Yggdrasil.

 

[amirm]

Then there is this from the company principal:

The company designers don't know how to measure their gear as well as some bloggers??? Is there any wonder then that their equipment measures poorly?

They now own a $28K analyzer and still don't know how to do this work? Did they just buy it for PR reasons? Here I thought they were going to learn to use it properly to improve their products.

 

[jtwrace]

Those measurements are made by someone else, using different instruments, and units without provenance and need to be confirmed. There definite flaws in how those tests are run. So please don't run with them until I get to confirm them.

If it wasn't such a POSchitt I'd buy one.

 

[Thomas Savage]

Not quite. I explained how my measurements were made to AP folks after pointing out to Jude that he was doing it wrong. They then worked with Jude on an implementation even though I had told Jude that the right solution on APx555 requires much more research. I have explained all of this to AP folks this morning again and issues in what Jude has published.

In a nutshell, they have set up a test where all the distortions and aberrations of the DAC are wiped clean. And they then declare: "oh look, it is linear down to -130 db" or whatever. Well duh. Of course if you remove all the noise and distortion from a DAC, it then looks accurate. Why bother running such a test when in real usage of the DAC no such filtering exists. Talk about running off with measurements with no thought of what they really mean and what benefit and correlation there is with audibility.

On top of that Jude continues to block all of my posts on head-fi. So no way for me to convey this information there. This is not the way we converge to a consensus.

So my advice remains: please wait to draw any conclusions until I remeasure the device with my APx555. Until then, my data remains 100% valid in pointing out serious issues in the performance of Yggdrasil.

I don’t understand why the folks at AP would setup the test with them if the results are functionally useless.

Why would they do that ? There’s more going on here.

It’s a real downer we can’t all just have a honest open conversation about these diffrent measurements.

 

[Ron Texas]

On top of that Jude continues to block all of my posts on head-fi. So no way for me to convey this information there. This is not the way we converge to a consensus.

People get banned or blocked from forums for all sorts of things, sometimes without warning I found. All it takes is being in opposition to whatever cool aid they are drinking.

 

[Thomas Savage]

People get banned or blocked from forums for all sorts of things, sometimes without warning I found. All it takes is being in opposition to whatever cool aid they are drinking.

It can be hard, you want a happy ship but at the same time a echo chamber holds zero integrity or real value beyond self indulgence and is nothing more than a safe house for those who can’t stand to be challenged.

Doing this job it’s really easy to see how places end up being so linear in thought and view , It’s just a matter of what’s being produced imo is it of value beyond pure self satisfaction, most places aren’t but you will be waiting a long time for members to thank you for enabling challenging discourse on ya forum I will say that much.

Admin often just want a easy life , keep work and stress to a minimum so a echo chamber suits that perfectly.

Then you have all the usual lord of the flys bullshit on top lol ......

Why do we do this...

 

[garbulky]

Not quite. I explained how my measurements were made to AP folks after pointing out to Jude that he was doing it wrong. They then worked with Jude on an implementation even though I had told Jude that the right solution on APx555 requires much more research. I have explained all of this to AP folks this morning again and issues in what Jude has published.

In a nutshell, they have set up a test where all the distortions and aberrations of the DAC are wiped clean. And they then declare: "oh look, it is linear down to -130 db" or whatever. Well duh. Of course if you remove all the noise and distortion from a DAC, it then looks accurate. Why bother running such a test when in real usage of the DAC no such filtering exists. Talk about running off with measurements with no thought of what they really mean and what benefit and correlation there is with audibility.

On top of that Jude continues to block all of my posts on head-fi. So no way for me to convey this information there. This is not the way we converge to a consensus.

So my advice remains: please wait to draw any conclusions until I remeasure the device with my APx555. Until then, my data remains 100% valid in pointing out serious issues in the performance of Yggdrasil.

Well, it sounds like you’re saying that the people from AP don’t know how to set up test to measure devices using their own machine when that is the purpose of the machine they built. But you appear to be the only one that does know how to. Maybe you do and they don’t. I certainly don’t know enough to say either way. But I am a little skeptical as to this as it stretches credibility. They after all are the ones making the machine. I would be more confident if I saw a post from Audio precision apologizing for their alleged mistake in helping Jude set up the audio precision test

 

[Blumlein 88]

Not quite. I explained how my measurements were made to AP folks after pointing out to Jude that he was doing it wrong. They then worked with Jude on an implementation even though I had told Jude that the right solution on APx555 requires much more research. I have explained all of this to AP folks this morning again and issues in what Jude has published.

In a nutshell, they have set up a test where all the distortions and aberrations of the DAC are wiped clean. And they then declare: "oh look, it is linear down to -130 db" or whatever. Well duh. Of course if you remove all the noise and distortion from a DAC, it then looks accurate. Why bother running such a test when in real usage of the DAC no such filtering exists. Talk about running off with measurements with no thought of what they really mean and what benefit and correlation there is with audibility.

On top of that Jude continues to block all of my posts on head-fi. So no way for me to convey this information there. This is not the way we converge to a consensus.

So my advice remains: please wait to draw any conclusions until I remeasure the device with my APx555. Until then, my data remains 100% valid in pointing out serious issues in the performance of Yggdrasil.

Could you explain a little bit about what they were doing to wipe out distortion and aberrations of the DAC.

 

[amirm]

Well, it sounds like you’re saying that the people from AP don’t know how to set up test to measure devices using their own machine when that is the purpose of the machine they built.

No, they do know how to "set up the machine." Question is, what are they being asked to set up?

What they were asked to help setup was detection of level while eliminating all distortion and noise. This is NOT what we want. If the device creates X amount of noise and distortion on top of Y signal, we want to measure both.

What they needed to ask instead was how to replicate this measurement I made on AP2522 but on APx555:

The test I ran takes into account distortion and noise and hence is able to differentiate between DACs easily. Theirs does not. I know because I replicated their method and it would no longer do anything useful.

But you appear to be the only one that does know how to.

I am currently the only one who:

1. Owns both the 2522 and APx555 analyzers
2. Have a large body of results and many DACs on my bench to evaluate using both analyzers

It has taken me good bit of effort to replicate the way the 2522 measured linearity on APx555 with the above tools.

The 2522 has a cascade of analog analyzer and digital analyzer. The APx555 is only digital. So there is no 1:1 relationship between how the two run.

 

[amirm]

Could you explain a little bit about what they were doing to wipe out distortion and aberrations of the DAC.

Sure.

The problem is challenging. They are attempting to measure linearity down to -140 dB. As you and I both know, there is no DAC in the world that produces meaningful signal at anything close to those levels. But importantly, there is no ADC in any analyzer that can do the same. Yet they tried anyway based ironically on advice I gave to AP . That if you use aggressive filtering of noise and distortion, you can indeed eliminate a lot of variability.

So they did that and took that the N'th degree. At -130 dB, the Schiit Yggdrasil DAC produces this according to Jude's measurement:

A jumbled up mess which proves what I said at the outset: that the output here is completely corrupt and does not at all represent any fidelity to the digital input the DAC was told to produce.

Jude runs this through an aggressive high-Q filter and gets this:

We see that the filter has completely removed all traces of distortion and noise. So of course if you then measure this, it shows that the DAC is doing well.

But that is NOT what we hear out of the DAC. Nor what it electrically produced. We are cleaning up the output of the DAC and then measure it, then declare it a winner.

The "trick" here is to use only the filtering necessarily for the ADC to not have its noise and distortion profile be below that of the DAC under test. This can only be done through a bunch of trial and error which I went through on APx555 analyzer. My older 2522 "happened" to do this well out of box. I tried many things including changing the excitation signal, settling parameters for measurements, custom filtering, etc. I finally found something that while may not be identical to 2522, is very comparable.

Summary
Any filtering in the analyzer cleans both the DAC and ADC output. It is tempting to select an exceptionally narrow filter to get rid of all noise and distortion as to even show accurate values to -140 dB. But we know such data is fictitious as we don't know how to build such DACs. By carefully selecting the filtering and analyzer setting however, we can get reasonable results to about -120 dB. Any attempt to go beyond that in my testing will lead one into a ditch.

P.S. The FFT method is even a more extreme case of such filtering as there, you get to look at one individual spike and ignore all other noise and distortion characteristics.

 

[amirm]

Let's remember that there are other measurement show the deficiencies of the Yggdrasil outside of the tricky linearity test such as this residual noise and distortion of a 1 kHz tone (itself filtered):

It is clear that the Schiit Yggdrasil is creating a ton of non-harmonic, random and otherwise spikes at low levels. These are what conspire to destroy its low level linearity. In other words, our measurements are self-consistent and all paint the same picture.

BTW, it is highly unlikely that the above distortion magically goes away in balanced output.

 

[amirm]

But I am a little skeptical as to this as it stretches credibility.

Why are you not skeptical of the motivation on behalf of Jude? You think he is going into this with just a search for the truth? If so, why ban me from responding to him? Or refusing to send me his AP project files?

There is only one truth here: he wants to help Schiit by invaliding my test results. Which is fine. But he needs to replicate my tests, not invent his own and say it still represents the same thing.

I would be more confident if I saw a post from Audio precision apologizing for their alleged mistake in helping Jude set up the audio precision test

AP has done nothing wrong to apologize for. They have spent time and effort helping Jude do something he should have known at the start, and had the benefit of my explanation here. If Jude had asked AP to help him replicate my 2522, they would have given him different advice (although not clear they would have been able to give him ultimately what he wants).

Ultimately there is a great "appeal to authority" both in your post and Jude's. That is not proper. The data has to stand on its two feet. Alternatively I can pull rank on Jude and even AP folks on my understanding of such topic. But I am not. So let's move on such tactics.

 

[rebbiputzmaker]

But that is NOT what we hear out of the DAC. Nor what it electrically produced. We are cleaning up the output of the DAC and then measure it, then declare it a winner.

Are you saying "not what we hear" at -120-140? But we cannot hear at this level, how is this an issue, and how is his filtering wrong if this is still well below audibility?

How much of what you are seeing is the item under test or the residual of the AP, or a combination? If you want to extrapolate that his filtering (clean signal shown) takes you above this unreasonable low level, so be it. This should all be extremely under the threshold of audibility. At some point IMO we have to decide are we testing audio equipment or just playing with numbers? YMMV

 

[rebbiputzmaker]

Alternatively I can pull rank

 

[amirm]

Are you saying "not what we hear" at -120-140? But we cannot hear at this level, how is this an issue, and how is his filtering wrong if this is still well below audibility?

Our hearing at best has around 116 dBSPL of dynamic range. So we don't to measure beyond this level of dynamic range and hence the reason my linearity measurements stop at -120 dB. Jude on the other hand going to -140 dB which as I said, not only makes no sense, but leads the testing to be incorrect in general.

 

[rebbiputzmaker]

Our hearing at best has around 116 dBSPL of dynamic range. So we don't to measure beyond this level of dynamic range and hence the reason my linearity measurements stop at -120 dB. Jude on the other hand going to -140 dB which as I said, not only makes no sense, but leads the testing to be incorrect in general.

Ok I see.

 

[Blumlein 88]

Sure.

The problem is challenging. They are attempting to measure linearity down to -140 dB. As you and I both know, there is no DAC in the world that produces meaningful signal at anything close to those levels. But importantly, there is no ADC in any analyzer that can do the same. Yet they tried anyway based ironically on advice I gave to AP . That if you use aggressive filtering of noise and distortion, you can indeed eliminate a lot of variability.

So they did that and took that the N'th degree. At -130 dB, the Schiit Yggdrasil DAC produces this according to Jude's measurement:

View attachment 13295

A jumbled up mess which proves what I said at the outset: that the output here is completely corrupt and does not at all represent any fidelity to the digital input the DAC was told to produce.

Jude runs this through an aggressive high-Q filter and gets this:
View attachment 13297

We see that the filter has completely removed all traces of distortion and noise. So of course if you then measure this, it shows that the DAC is doing well.

But that is NOT what we hear out of the DAC. Nor what it electrically produced. We are cleaning up the output of the DAC and then measure it, then declare it a winner.

The "trick" here is to use only the filtering necessarily for the ADC to not have its noise and distortion profile be below that of the DAC under test. This can only be done through a bunch of trial and error which I went through on APx555 analyzer. My older 2522 "happened" to do this well out of box. I tried many things including changing the excitation signal, settling parameters for measurements, custom filtering, etc. I finally found something that while may not be identical to 2522, is very comparable.

Summary
Any filtering in the analyzer cleans both the DAC and ADC output. It is tempting to select an exceptionally narrow filter to get rid of all noise and distortion as to even show accurate values to -140 dB. But we know such data is fictitious as we don't know how to build such DACs. By carefully selecting the filtering and analyzer setting however, we can get reasonable results to about -120 dB. Any attempt to go beyond that in my testing will lead one into a ditch.

P.S. The FFT method is even a more extreme case of such filtering as there, you get to look at one individual spike and ignore all other noise and distortion characteristics.

Well that explains why my linearity results have been better than expected.

I didn't have any automated test gear or software that would do the job for me. So just thinking about a good way I decided to use a quarter sample rate tone that way you get one positive bit, zero, and one negative bit turned on. Play a few seconds of that and step down one bit in level for a few more seconds. When done measure the level of the tone in an FFT, and pretty much all the things that weren't junk were good until the noise of each FFT bin gets close to the bits for the tone. Also I would put a 1 khz tone at -60 db in the beginning. Use gain on the input of the ADC so the noise floor of the DAC under test was raised well above that of my ADC. Then digitally adjust level back down to -60 db so everything was where it should be level wise.

I'm not sure that is a bad way to do it. The point of the test in my mind was to see if the device puts out the signal at the right level for each bit. The DAC could do this and a poor analog section or poor clocking could obscure the result. I'm doing other tests for noise level, jitter, and distortion so I didn't want these rolled into linearity of the DAC chip in use. Or at least that is how I was thinking about it. It is true we don't get to hear the signal without those, but that wasn't the point of this test. I'm not going to listen to level descending 12 khz tones any way.

So what sort of signal is the AP using for this? Seems I recall a quick way to check it using a search algorithm and a least squares fit. I suppose that is to give a result for all 65K possible levels instead of just the 24 bits with one bit turned on at a time.

 

[DonH56]

https://standards.ieee.org/findstds/standard/1241-2010.html -- for ADCs but the methodology works for DACs

This one is for DACs; I don't have it (I helped with the ADC Standard, no involvement with this one): http://standards.ieee.org/findstds/standard/1658-2011.html

My copy of 1241 is an older version; I should pony up and get them both, but am lazy and broke.

 

[Thomas Savage]

When we test linearity (and indeed every test we do here) don’t we want to know how the device under test would perform under normal use (in this case) in your home?

Unless your just wanting to play number top trumps for marketing reasons of course, but as for ‘consumer advice’ amir is doing what’s pertinent to the practical application of these devices.

We are all here for consumer advice to one degree or another.

Please let’s not go after AP here, they are in a impossible position really and it’s not for them to get involved in this ( not publicly anyway) .

 

[mindbomb]

It is clear that the Schiit Yggdrasil is creating a ton of non-harmonic, random and otherwise spikes at low levels.

Wait, I thought the analog v2 upgrade fixed those issues, as it looks like on the 1khz non-filtered thd here. Or is it just easier to see with the 1khz filtered out?