Review and Measurements of Schiit Yggdrasil V2 DAC

[Alcophone]

Have you seen the infamous video, Digital Show and Tell by Monty Montgomery?

Those all analog HP high precision analyzers sure do seem to find the digital wave plenty smooth.

Nice video, thank you! I've seen a different video of his before, but not that one.

@amirm should watch it, too, starting at 15:07
"But dither isn't only an on or off choice. We can reduce the dither's power to balance less noise against a bit of distortion to minimize the overall effect."

Slightly related to Monty's video, I liked this video, too:

 

[tranq]

I saw this posted on another thread I participate on.

Im not going to worry much about measurements anymore.

Yes, I can see how this video is open to interpretation/discussion. I'm going to take the side that pleases me.

We all should do the same, without prejudice. Life's too short to get worked up over how folks choose DACs.

 

[amirm]

I saw this posted on another thread I participate on.

Im not going to worry much about measurements anymore.

Yes, I can see how this video is open to interpretation/discussion. I'm going to take the side that pleases me.

We all should do the same, without prejudice. Life's too short to get worked up over how folks choose DACs.

Paul is a lovely person to listen to. I often watch his videos. Half the time he has something informative to say. The other times he is completely wrong. But still entertaining.

If you read my tutorial on audio measurements, https://www.audiosciencereview.com/forum/index.php?threads/understanding-audio-measurements.2351/, you will see this:

Perceptually THD+N is a rather poor metric since it considers all harmonic distortion products to have the same demerit. As I explained, from masking point of view, later spikes are more important than earlier ones. So don't get fixated on small differentials between devices being tested. Large differences like what is shown above though are significant and show the difference between good design/engineering and not so good ones.

With respect to Schiit Yggdrasil, we are dealing with the bolded section. In other words, measurements are there to show you the level of competence and quality of engineering. If that is poor, at best that DAC will sound the same as any other DAC. At worst, it will sound worse. In no case is there a magic component that makes it sound better.

Let's remember that the DAC has all but one purpose: convert digital samples faithfully to analog. The more it spits on the samples, the worst it is.

As to Paul's advice to use your ears, sure, you can do that. You would need access to the DACs of course and ability to match levels. And perform the test without bias. These are factors that Paul won't tell you about because he wants you to buy more expensive DACs based on emotional need or perception, rather than pure sound quality.

Ultimately remember that as nice as Paul comes across as a person, he is a manufacturer and likes to sell you gear. And that selling in the case of DACs, relies on differentiating his product from commodity ones. He can't do it by measurements so he sends you down the subjective, faulty listening test results.

Of course you are welcome to ignore all of this and go by what you like. But before you do, google "carrot juice cure cancer." I did that once randomly and as I guessed, there were countless people claiming they have completely cured their cancer by drinking carrot juice. If you dismiss that out of hand, then please do the same for audio. Do not believe people like Paul. Believe the proper audio science which is what we advocate and explain here.

 

[amirm]

@amirm should watch it, too, starting at 15:07
"But dither isn't only an on or off choice. We can reduce the dither's power to balance less noise against a bit of distortion to minimize the overall effect."

You can shape that noise and match it to threshold of hearing. In the case of Yggdrasil that is not even required since the noise floor we are talking about is with respect to 20 bit samples which is below threshold of hearing. There is no reason to leave quantization noise there.

 

[rebbiputzmaker]

because he wants you to buy more expensive DACs based on emotional need or perception, rather than pure sound quality.

A question:

Why is it always buy" the most expensive", why not if a person uses common sense would they not buy the least expensive item that meets their criteria? Even if they use flawed listening tests, why would it always be the most expensive item winning? Most people shop for the "best for less" trying to get value for their hard earned money. Why is this necessarily different in this case?

 

[Frank Dernie]

A question:

Why is it always buy" the most expensive", why not if a person uses common sense would they not buy the least expensive item that meets their criteria? Even if they use flawed listening tests, why would it always be the most expensive item winning? Most people shop for the "best for less" trying to get value for their hard earned money. Why is this necessarily different in this case?

This has not been my experience.
In the hifi business everything is orientated, either directly or subliminally, at the notion that more expensive is better and if there is a comparison it is almost always between items of the same price, I suppose to avoid the embarassing fact that an inexpensive item may well be audibly indistinguishable from another at 10% of its price. I had this experience when auditioning DACs myself. The expensive one was in a much better looking machined from solid ally case and had a remote whereas the one at a tenth of the price was in a painted steel case and no remote but the sound was the same (both excellent).
The alternative reason could because manufacturers and dealers want us to be on the upgrade bandwagon, and that only works on people convinced that more expensive must be better.
I have not had this experience with speakers, tape recorders or record players!

 

[rebbiputzmaker]

This has not been my experience.
In the hifi business everything is orientated, either directly or subliminally, at the notion that more expensive is better and if there is a comparison it is almost always between items of the same price, I suppose to avoid the embarassing fact that an inexpensive item may well be audibly indistinguishable from another at 10% of its price. I had this experience when auditioning DACs myself. The expensive one was in a much better looking machined from solid ally case and had a remote whereas the one at a tenth of the price was in a painted steel case and no remote but the sound was the same (both excellent).
The alternative reason could because manufacturers and dealers want us to be on the upgrade bandwagon, and that only works on people convinced that more expensive must be better.
I have not had this experience with speakers, tape recorders or record players!

That's fair, but in the end, it is still the person's decision. Even armed with all the data that might steer them otherwise, they could still make that impulsive purchase. Pretty hard to get a person with an addictive personality to give up their addiction. People often lusting for that fix.

 

[Blumlein 88]

Obversely, why doesn't high THD sound bad. The answer is if you get THD high enough, it will sound bad.

Now once you get THD low enough, less doesn't matter. It does in fact equate to better engineering. In between is a zone where some might prefer a certain amount of certain harmonics where more is worse, and less is worse. Damn those humans, they always muck up the simple good story.

 

[amirm]

That's fair, but in the end, it is still the person's decision. Even armed with all the data that might steer them otherwise, they could still make that impulsive purchase. Pretty hard to get a person with an addictive personality to give up their addiction. People often lusting for that fix.

There is nothing impulsive about buying a very expensive piece of hifi. People are conditioned to believe and that is what they purchase.

But maybe you tell us that there is no reason for education of any kind.

 

[rebbiputzmaker]

Obversely, why doesn't high THD sound bad. The answer is if you get THD high enough, it will sound bad.

Now once you get THD low enough, less doesn't matter. It does in fact equate to better engineering. In between is a zone where some might prefer a certain amount of certain harmonics where more is worse, and less is worse. Damn those humans, they always muck up the simple good story.

Ok maybe like too much salt will raise your blood pressure if you are hypertensive, too little the food will be tasteless, but there is a level that will still keep you out of the ER but will satisfy your taste.

 

[rebbiputzmaker]

But maybe you tell us that there is no reason for education of any kind.

Education is important, useful and appreciated, but in some cases, human will defy all logic and just be themselves.

 

[RayDunzl]

Im not going to worry much about measurements anymore.

Music is distortion.

A little more (maybe, measureably, a lot more) isn't easily noticed (particularly without a clean reference with which to immediately compare).

Here's a single bass guitar note. 22.7% THD on the fundamental.

 

[solderdude]

Education is important, useful and appreciated, but in some cases, human will defy all logic and just be themselves.

You mean in some cases humans are just being stubborn or refuse to accept something, based on their rather limited, flawed and easily fooled personal perception (ears, sight, smell taste, touch) and ways of reasoning and drawing conclusions ?

 

[RayDunzl]

If you really want undistorted music, the closest thing I've found (that isn't electronic) would be a group that gently blows on glass bottles...

Grolsch 330ml Beer Bottle. 0.4% THD

 

[solderdude]

How could one be certain the small distortion products seen are even there in reality ?
Maybe the mic or its pre-amp add the harmonics ?

 

[DuxServit]

The core notion (which apply to most consumer products) is to equate value — perceived or true — to certain goods, even if there wasn’t one to begin with.

A good example is bottled water, specifically in France which has good tap water. Prior to the 1950s, water was a non-mention. Then a couple of corporations begun associating value to bottled water through extensive marketing (e.g. helps indigestion; makes better skin for women; slimming; etc).

Today there are over 30 different brands of water sold in France with very different price points. At the low end a bottle can sell for Euro 0.20. At the high end food stores, some bottles sell for Euro 7.00.

Seems audio gear shares a lot of marketing characteristics as bottled water.

 

[Wombat]

Ok maybe like too much salt will raise your blood pressure if you are hypertensive, too little the food will be tasteless, but there is a level that will still keep you out of the ER but will satisfy your taste.

Tried the Pritikin Diet once-upon-a-time(wife's idea). Once weaned off salt, sugar and fat I enjoyed, and still do, food's natural flavors without those additives.

 

[mindbomb]

Im not going to worry much about measurements anymore.

well, I'm sure you can save a lot of money then and just stick with the built in dacs in devices.

 

[andreasmaaan]

I saw this posted on another thread I participate on.

Im not going to worry much about measurements anymore.

Yes, I can see how this video is open to interpretation/discussion. I'm going to take the side that pleases me.

We all should do the same, without prejudice. Life's too short to get worked up over how folks choose DACs.

THD is not a good metric, and does not correlate well to perceived sound quality.

But @amirm does not only measure THD, he measures and graphs the specific harmonics produced by a device at a particular frequency.

There are various distortion metrics other than THD that correlate far better with perceived sound quality (examples: Gm, R(nonlin), etc.). These unfortunately involve more complex maths, which is why I believe THD is often used (and also ofc because it can make a product look better on paper than it really is).

These superior metrics other than THD are based (among other things) on the general principle, which is strongly supported by experimental data, that lower order harmonics will be masked much more effectively by the fundamental than higher order harmonics when processed by the human auditory system.

So, even without using the complex maths required by these superior metrics, we can simply apply this general principle to Amir's graphs and look at the levels of the various harmonics to estimate their audibility (or otherwise).

For example, a 2nd harmonic at -70dB is far less likely to be audible than a 9th harmonic at -70dB, because the 2nd harmonic is more likely to be masked (it is closer in frequency to the fundamental). However, both 2nd and 9th harmonics are weighted evenly when calculating THD, and this is why THD is of very limited usefulness in questions of sound quality.

The lesson from this is not that we should abandon measurements entirely, but that we should understand a particular measurement's powers and its limitations. And that we should not necessarily settle for the simplest measurement when interpreting how a particular device will perform sonically. Fortunately there are many far more powerful and useful measurements than THD available.

 

[Krunok]

THD is not a good metric, and does not correlate well to perceived sound quality.

There are various distortion metrics other than THD that correlate far better with perceived sound quality (examples: Gm, R(nonlin), etc.). These unfortunately involve more complex maths, which is why I believe THD is often used (and also ofc because it can make a product look better on paper than it really is).

These superior metrics are based among other things on the general principle, which is supported by experimental data, that lower order harmonics will be masked much more effectively by the fundamental than higher order harmonics.

Of what practical value are those distortion metrics if new generation measuring devices, like the one @amirm uses, are not able to support them?