• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Measurements and Review of Schiit Yggdrasil DAC

Status
Not open for further replies.

GearMe

Member
Joined
Feb 7, 2018
Messages
36
Likes
16
@GearMe There’s the data from Amirs testing here but also a treasure trove of information on countless topics related to audio and a deal of well imformed members to talk about it with.

To turn data into useful information takes a deal of understanding but there’s help here for that too. ( don’t ask me though , I just work here :D)

Thanks!

Let me address this since we are being stereotyped as being all about measurements. If you look more broadly in this forum, you will see extensive discussions of audio science, acoustics, audio theory, etc. So we are about totality of audio science and engineering.

On the specifics of my measurements, they for the most part indicate engineering excellence or lack thereof. Where appropriate I have and continue to make comments about audibility. Many times I dismiss the visual differences as not being audible differences. No one should jump to the conclusion that the measurements indicate audible superiority by the scores seen in the graphs. For good or bad, we are really, really bad at hearing these unfamiliar distortions no matter how much we think otherwise.

In the context of this review, please take a look at frequency response error in the upgraded analog board. It shows a 1 dB drop at low frequencies. I cited research that shows that is audible and is an audible problem. That is a case of graphs matching audible differences. Schiit needs to take immediate action to investigate this problem. They need to measure their own boards, reach out to that owner, etc. to figure out what is going on.

Yes...this is helpful! Much like the info in your 'Audibility of Small Distortions' article is. Given your background/capabilities, to the degree that you could elaborate on the audibility within the context of a particular review, it would be even more helpful!

An example of this would be Tyll's use of the combined measurements and comments page at the end of each of his reviews....link below is to a recent one. Not expecting you'd want to do a full-on subjective review but possibly something like this?
https://www.innerfidelity.com/conte...r-ear-planar-magnetic-headphones-measurements

His comments/interpretations can be tied back to the Headphone Measurements Explained tab where you can go to refresh your memory on interpreting/aligning measurement differences between cans

The example pages below are related to Interpreting Square Wave Responses
https://www.innerfidelity.com/content/headphone-measurements-explained-square-wave-response
https://www.innerfidelity.com/content/headphone-measurements-explained-square-wave-response-page-2

Re: the measured Gen 2 board dip, is it a 1 dB drop or a .5 dB drop? Want to make sure I'm reading these things right!

FWIW - on Schiit needing to take immediate action, that's on them obviously and the owner if he/she complains. I suppose if enough of their customers feel this is an issue, then that will happen.

One note though, I'm guessing that many of the Yggy owners use balanced mode and, therefore, wouldn't be experiencing this issue...just a guess.
 

maul

Member
Joined
Nov 1, 2017
Messages
15
Likes
29
I do not agree with your assertion though and I do not believe the measurements Amir has posted of anything I own are even bad measurements for the equipment.

It sounds like you're just in straight up denial at this point, because there were issues pointed out even in the Stereophile review. I would like to hear an explanation for how these measurements don't reveal issues however, if you want to oblige.
 

Thomas savage

Grand Contributor
The Watchman
Forum Donor
Joined
Feb 24, 2016
Messages
10,260
Likes
16,298
Location
uk, taunton
It sounds like you're just in straight up denial at this point, because there were issues pointed out even in the Stereophile review. I would like to hear an explanation for how these measurements don't reveal issues however.
It does not sound like that to me, it sounds like a guy who’s used to a certain presentation ( maybe in their day to day profession) vs a presentation aimed at guys skipping through audio forums .

Personally I’d like to see a parallel thread to each review with the measurement setup, a few pictures and have it open for discussion. That solves both our aims to keep things ‘ digestible ‘ and the interests of those like this gentleman ( @rtg97229 )
 

maul

Member
Joined
Nov 1, 2017
Messages
15
Likes
29
Well, I'm eagerly waiting an explanation beyond trying to discredit the equipment/method. Because on one hand Mike Moffat is saying it's horrible linearity and the Bifrost he used, for instance, must be broken... while at the same time atomicbob's measurements show the exact same thing for that particular test, and Jason is now saying that r2r measures differently. So which is it? I'd also like to see some alternate measurements, maybe by Schiit themselves even...
 

Thomas savage

Grand Contributor
The Watchman
Forum Donor
Joined
Feb 24, 2016
Messages
10,260
Likes
16,298
Location
uk, taunton
Well, I'm eagerly waiting an explanation beyond trying to discredit the equipment/method. Because on one hand Mike Moffat is saying it's horrible linearity and the Bifrost, for instance, must be broken... while at the same time atomicbob's measurements show the exact same thing for that particular test, and Jason is now saying that r2r measures differently. So which is it? I'd also like to see some alternate measurements, maybe by Schiit themselves even...
Oh yea no doubt, Iv no issue that the data from Amirs tests are a true reflection of the performance of the unit(s) he tested.
 

rtg97229

Member
Joined
Feb 16, 2018
Messages
19
Likes
7
It sounds like you're just in straight up denial at this point, because there were issues pointed out even in the Stereophile review. I would like to hear an explanation for how these measurements don't reveal issues however, if you want to oblige.

I think you are referring to something different in the Stereophile measurements than what is measured here. I made all of my Schiit purchases after that review was already out and I don't remember all of the results so it was not a large factor in my decision. If you are talking about the 0 crossing glitch that is an issue with the R2R DACs before Schiit (I do not like the name personally) started making DACs. I do have some questions for Schiit about their solution to the glitch and the tradeoffs but I don't expect them to answer those here. I did not see it in Amir's measurements so they where honest about addressing the issue at least. I also own a Freya that is likely to measure poorly compared to other preamps that I own. I still enjoy it but that does not mean measurements are not also interesting.
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,385
Location
Seattle Area
I think you are referring to something different in the Stereophile measurements than what is measured here. I made all of my Schiit purchases after that review was already out and I don't remember all of the results so it was not a large factor in my decision.
As a test and measurement guy, what data then led to your purchase of Schiit products?
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,385
Location
Seattle Area
If you are talking about the 0 crossing glitch that is an issue with the R2R DACs before Schiit (I do not like the name personally) started making DACs.
Not at all. The DAC chip Analog Devices offers which is used in Schiit multi-bit DACs is designed for domains outside of audio. To use it for audio, you need to add further measures to it, the top of which is accomodating glitching. Other companies have for example used parallel DACs and interleaved between them allowing the data to settle before switching.

Here is another scheme discussed about another ladder DAC by someone who is well aware of this problem:
I have a little tech info at http://www.soekris.dk/products.html, or maybe this snippet from the PCM1704 datasheet help:

The sign-magnitude architecture, which steps away from
zero with small steps in both directions, avoids any glitching
or large linearity errors, and provides an absolute current
output. The low-level performance of the PCM1704 is such
that true 24-bit resolution can be realized around the critical
bipolar zero point.

As I said, Sign Magnitude can be a little tricky to comprehend, don't know how to explain if further.

BTW, the problem is not "zero switching." That is where it shows up the most. The problem is that you can't assure all the switchers in the ladder DAC activate at the same time. This accuracy error exists all the time. It just shows up the most at zero crossing because all of them are switching state.

By letting manufacturers get away with not publishing any measurements, you let them get away with poor design decisions.
 

rtg97229

Member
Joined
Feb 16, 2018
Messages
19
Likes
7
As a test and measurement guy, what data then led to your purchase of Schiit products?

Oh price point was number one on the list fore sure. Trying something different was also a factor. My doubts about the company are mostly concerns about long term quality and customer service. I both like the idea of internet direct and have concerns about it. I am not saying that measurements are not interesting for audio equipment, they are and I am happy people are taking the measurements.
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,385
Location
Seattle Area
You are again taking my words out of context needlessly in a conversation that could have easily been done as soon as we agreed that I am not your target audience.
That is a flippant comment meant again to spread FUD. I have asked you repeatedly for constructive suggestions on what you think is wrong with my measurements and you are not forthcoming.

I was very specific with you over on head-fi and you did not respond. Here you have been a bit better but strangely defensive given that my questions are not difficult to either answer or simply state that it is not something that you believe to be interesting to your target audience.
Kind of hard to respond when I was banned from that thread. Here is your post from there:

upload_2018-2-17_12-26-53.png


I have provided this documentation: https://www.audiosciencereview.com/forum/index.php?threads/understanding-audio-measurements.2351/

It is in a single place and not only explains the test methodology but also includes a tutorial on what the test does and examples of what it means.

To which you responded with this:
Was just trying to help, if you don't think you need it that is fine. I work in test and measurement for a living and if someone handed me a calibration report on one of my SAs with your level of documentation I would drop it in the garbage.

I have since asked you who provides calibration for audio reviews and you refuse to answer. I will answer that for you: no one. Simple reason is that absolute accuracy is not necessary for the message we try to convey.

You asked for picture of the setup. I have provided one in my link above. Beyond that I asked you what good would come out of showing a few audio cables going to and from the DAC and you refuse to explain.

There is no such pictures in John Atkinson's measurements. Are you on record saying his measurements should also be thrown out?

You talk about test modes. That is detailed in the above link. Have you read it?

Test date? Honestly? You didn't get a sense of when I tested the gear from this review? And you would throw out the results for that?

Bottom line is that you don't like the outcome and don't want the people to have the data. So you spread doubt about it. It is improper, non-constructive and does a disservice to the community at large.
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,385
Location
Seattle Area
Oh price point was number one on the list fore sure. Trying something different was also a factor.
Oh, as a test and measurement guy you did not seek such information and went by price point? How did you put the value of your profession aside in making such an expensive purchase decision?

And are you not aware of much cheaper alternatives?

What was different here? How did you ascertain that? Any hope of measurement data being behind that?
 

rtg97229

Member
Joined
Feb 16, 2018
Messages
19
Likes
7
Not at all. The DAC chip Analog Devices offers which is used in Schiit multi-bit DACs is designed for domains outside of audio. To use it for audio, you need to add further measures to it, the top of which is accomodating glitching. Other companies have for example used parallel DACs and interleaved between them allowing the data to settle before switching.

Here is another scheme discussed about another ladder DAC by someone who is well aware of this problem:


BTW, the problem is not "zero switching." That is where it shows up the most. The problem is that you can't assure all the switchers in the ladder DAC activate at the same time. This accuracy error exists all the time. It just shows up the most at zero crossing because all of them are switching state.

By letting manufacturers get away with not publishing any measurements, you let them get away with poor design decisions.

Delay time for switching in a R2R is an interesting topic and part of what I think Schiit should be asked about given that their statement about fixing the glitch was that it was solved in DSP. I don't yet understand how DSP addresses your concern. I don't know that I would call what they have made a poor design but it is perfectly fair to ask questions about design compromises.
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,385
Location
Seattle Area
An example of this would be Tyll's use of the combined measurements and comments page at the end of each of his reviews....link below is to a recent one. Not expecting you'd want to do a full-on subjective review but possibly something like this?
https://www.innerfidelity.com/conte...r-ear-planar-magnetic-headphones-measurements
I can't do what he does. His subjective remarks are non-scientific because they are not in controlled blind tests with level matching. He says this for example:

"Bass is ruler flat to 10Hz; a look at the 30Hz square wave and virtually non-existan bass distortion validate my experience that the LCD2C has fantastic bass response."

Who says it has fantastic bass response? Just because he says it, doesn't make it true.

And where is the science behind 30 Hz square wave having any predictive power here with respect to audibility?

He is mixing objective measurements with subjective non-scientific remarks. In this forum, we are not part vegetarian that way. :) We always follow the science and I cannot and will not throw in subjective remarks that are not created based on proper testing.

Where appropriate based on psychoacoustics I do make audibility remarks but that is the extent of what I can do.
 

rtg97229

Member
Joined
Feb 16, 2018
Messages
19
Likes
7
That is a flippant comment meant again to spread FUD. I have asked you repeatedly for constructive suggestions on what you think is wrong with my measurements and you are not forthcoming.

I disagree, I thought we had an understanding on this topic already.

Kind of hard to respond when I was banned from that thread. Here is your post from there:

View attachment 10697

I have provided this documentation: https://www.audiosciencereview.com/forum/index.php?threads/understanding-audio-measurements.2351/

It is in a single place and not only explains the test methodology but also includes a tutorial on what the test does and examples of what it means.

I did not know when you got ban and I do not like that they censored you. I should look back at what I wrote because I think I thanked you for what information you did provide. It is not the format I am used to but you did respond that you are targeting a different audience and I have no problem accepting that answer.

Test date? Honestly? You didn't get a sense of when I tested the gear from this review? And you would throw out the results for that?

Normal information for reports that I am used to reading. This has already been covered and I am starting to feel bad for the dead horse.

Bottom line is that you don't like the outcome and don't want the people to have the data. So you spread doubt about it. It is improper, non-constructive and does a disservice to the community at large.

Simply not true and I do not understand how you come to this conclusion. Much like I agree with you or anyone else asking questions of a manufacture I also see no problems with asking you questions.
 

garbulky

Major Contributor
Joined
Feb 14, 2018
Messages
1,510
Likes
827
It does not sound like that to me, it sounds like a guy who’s used to a certain presentation ( maybe in their day to day profession) vs a presentation aimed at guys skipping through audio forums .

Personally I’d like to see a parallel thread to each review with the measurement setup, a few pictures and have it open for discussion. That solves both our aims to keep things ‘ digestible ‘ and the interests of those like this gentleman ( @rtg97229 )
Sorry I posted without reading what you said. Yes this is what I was talking about which you worded in a much better way.
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,385
Location
Seattle Area
So how about a thread called boring measurements. Then you post your review using all your normal brevity and do it just the way you like. After that you link to the related post in "boring measurements and test setup details" where you can find all the baseline stuff that you regard to be just run of the mill stuff that nobody is interested in. That way if people say this is invalid without certain setup data, they can suggest to you any data they are finding to be missing and you can stick it all there in the thread that won't interfere with your actual review.
Thanks for your post. I appreciate the points within. Two comments:

1. I am not hearing criticism from anyone genuinely interested in additional information. Every comment comes from people with obvious angst about the data as presented. Without fail, none have explained what the additional data would do to improve what is presented. Nor can they cite any precedence for such documentation. Indeed many happily accepted other measurements lacking such (e.g. calibration certs).

2. I don't get paid for any of this. :) Seriously this is a hobby for me. I like measuring gear and seeing how they perform. Writing a dissertation, taking extra pictures to resize, upload, etc. is just busy work that takes away the fun and importantly, time that I have that I can put toward testing more gear.

I already created that thread on understanding audio measurements and put fair amount of work in it. Despite that, the people who demanded such, still are making more demands. While I am committed to improving that write-up, I am just not seeing that as a good use of my time, or value to readers.

Folks can always ask me for additional data in my measurement and review threads and I am always willing to perform them. I was asked for example for warm-up testing and I did that for Bifrost. You want me to tell you when I measured the device? Just ask.

So it is not like people can't ask questions and get answered. I am just objecting to rewarding their PR attempts to diffuse data by creating work for myself. How about them going and doing something useful like asking manufacturers to publish measurements?
 

Dimitri

Senior Member
Joined
Jan 28, 2018
Messages
365
Likes
426
Location
Valencia California

For someone who is more impressed by documentation than credentials, should you be knocking a little harder on the "company in question"' door to provide their own test reults for their product?


I don't know that I would call what they have made a poor design but it is perfectly fair to ask questions about design compromises.
There is a fine line between a poor design and a design compromise. Either one can be justified by the designer, who seems to be in no mood to get involved in such explanation.
Yet you believe it's fair to ask.
Go ahead, no one here is holding you back :) Worse thing that can happen is you'll end up banned from HF as well.
 

Thomas savage

Grand Contributor
The Watchman
Forum Donor
Joined
Feb 24, 2016
Messages
10,260
Likes
16,298
Location
uk, taunton
Sorry I posted without reading what you said. Yes this is what I was talking about which you worded in a much better way.
The measurements themselves are done within a very simple setup though, the conditions being the same for every unit measured.
 

Dimitri

Senior Member
Joined
Jan 28, 2018
Messages
365
Likes
426
Location
Valencia California
I am just objecting to rewarding their PR attempts to diffuse data by creating work for myself.
Have you considered banning people and deleting their posts? I've heard of "post cleansing" happening in other forums.
j/k :)
 
Status
Not open for further replies.
Top Bottom