Measurements and Review of Schiit Yggdrasil DAC

[FrivolsListener]

I own a recently purchased Yggdrasil and I'm not happy. I wrote a message to the support team to ask them to please explain to me why the testing shows it cannot even process 16 bit adequately. I cannot wait to see if I get a response and if so what it says. I have wasted enough money on the voodoo science of cables and this I find potentially much worse. I feel a google review brewing.

First off, hold on. Don't be "That Guy." It took me half a lifetime to learn to not go after companies like that. You end up pissing everyone off, which helps no one.

Second, this site is a single data point. Question it and look for others. Question those, too. For those who participate in forums like this and elsewhere, ask what their results mean. Hell, go to head-fi and ask Baldr what the measurements mean. Ask atomicbob on SBAF what they mean. Contact the reviewers at Stereophile and TAS and ask them, too. Take no one's word as gospel, but hear everyone out.

 

[March Audio]

First off, hold on. Don't be "That Guy." It took me half a lifetime to learn to not go after companies like that. You end up pissing everyone off, which helps no one.

Second, this site is a single data point. Question it and look for others. Question those, too. For those who participate in forums like this and elsewhere, ask what their results mean. Hell, go to head-fi and ask Baldr what the measurements mean. Ask atomicbob on SBAF what they mean. Contact the reviewers at Stereophile and TAS and ask them, too. Take no one's word as gospel, but hear everyone out.

You mean pissing schitt off.

If the measurements are wrong, and I dont think they are, schitt can produce measurement evidence of their own which demonstrates the contrary. This should be a trivial exercise as Im sure they measured the performance during development

The only validity to your position here is that I beleive that if Jerry is disatisfied with the product his energy would be better spent selling it and moving on. Arguing with schitt is unlikely to produce anything useful imo.

 

[Jerry Sobel]

Your point is well taken but a measurement is a measurement and I just want the company to tell me what it means. I think after spending over 2K I deserve that. If I go onto other forums I will be run out of town. I do not know all the jargon etc.

 

[garbulky]

You are correct and I apologize for my part in it.

The problem is
Anyway, stick around, and perhaps we'll all learn something. - Don

Thank you. Maybe it doesn't help that I assess audio subjectively. But I am open minded. I like what Amir says, the measurement stands for itself. The intepretation of it is a different thing.

I think it's good that the audio world is be held to a higher standard. Many companies don't publish their measurements. Some won't even publish basic specs or even pictures of the inside of their gear. So at least some one is trying to do that.
Personally I hope there will come a day that subjective impressions can be quantified using measurements in greater detail than what is possible now.

 

[FrivolsListener]

Your point is well taken but a measurement is a measurement and I just want the company to tell me what it means. I think after spending over 2K I deserve that. If I go onto other forums I will be run out of town. I do not know all the jargon etc.

I totally disagree -- a measurement is not a measurement. One can lie with measurements easily by careful omission of facts or settings that are wrong for the type of measurement the instrument is doing, or with a poorly chosen instrument for the job. It's like lying with statistics.

That said, I don't think @amirm 's measurements are intentionally in error. But the discrepancy is going to drive me nuts for a while.

 

[FrivolsListener]

You mean pissing schitt off.

Or, you know, anyone who has to deal with the public. I have a rule; don't be the, "I want to speak to the manager," lady straight out of the gate. Whoever you are talking to is also a person and they are probably working in good faith. (At least, assume so at first.)

 

[Blumlein 88]

None of that means what you say (except the bit about it not being a virtue). It means this DAC takes longer to stabilize. This DAC is meant to be left turned on.

It could well be. Measurements could go a long way to showing that. Until then it's somebody arguing against the manufacturer.

I've provided measurements of some gear from dead cold to 2 hour warm up every 30 minutes and then after 24 hours. You can't convince people when the designer says otherwise and they think they hear it happen. Of course I can't say my measurements are applicable to every piece of gear made I didn't measure. Yet if you do this a few times and realize clocks stabilize in seconds with the only changes being in the less than ppm range after that while analog circuits may drift a wee little bit more yet still barely above the lowest we can measure then the onus for 72 hours is on the maker of said gear.

 

[Blumlein 88]

First off, hold on. Don't be "That Guy." It took me half a lifetime to learn to not go after companies like that. You end up pissing everyone off, which helps no one.

Second, this site is a single data point. Question it and look for others. Question those, too. For those who participate in forums like this and elsewhere, ask what their results mean. Hell, go to head-fi and ask Baldr what the measurements mean. Ask atomicbob on SBAF what they mean. Contact the reviewers at Stereophile and TAS and ask them, too. Take no one's word as gospel, but hear everyone out.

As a second data point here is JA's description after measuring an earlier version of the same device (from one year ago).

It's difficult to sum up the Schiit Yggdrasil's measured behavior. While the processor's analog circuitry is superbly well designed, its digital circuitry appears to have problems with high-level, high-frequency tones, and with the LSBs of 24-bit data. It's possible, of course, that the former will be rare with music, and that the latter will be obscured by the noise floors of recordings. But it does look as if the digital circuitry is not fully optimized. Hopefully, this could be addressed with a firmware upgrade.—John Atkinson
Read more at https://www.stereophile.com/content...da-processor-measurements#jrceVBYdWVT2v1yd.99

 

[Blumlein 88]

Thank you. Maybe it doesn't help that I assess audio subjectively. But I am open minded. I like what Amir says, the measurement stands for itself. The intepretation of it is a different thing.

I think it's good that the audio world is be held to a higher standard. Many companies don't publish their measurements. Some won't even publish basic specs or even pictures of the inside of their gear. So at least some one is trying to do that.
Personally I hope there will come a day that subjective impressions can be quantified using measurements in greater detail than what is possible now.

I think the problem with your last sentence above is much of the gear has surpassed what humans can discern. Humans don't like being told that. So they search for what isn't there or bump into things that really do sound different because of an artful coloration (and therefore necessarily of lower fidelity).

Some fair amount of academic research into human hearing maps out at least in broad outline if not better things that can and can't be heard and at what levels. Audiophiles just will not be convinced by it vs subjective experience.

 

[FrivolsListener]

As a second data point here is JA's description after measuring an earlier version of the same device (from one year ago).

It's difficult to sum up the Schiit Yggdrasil's measured behavior. While the processor's analog circuitry is superbly well designed, its digital circuitry appears to have problems with high-level, high-frequency tones, and with the LSBs of 24-bit data. It's possible, of course, that the former will be rare with music, and that the latter will be obscured by the noise floors of recordings. But it does look as if the digital circuitry is not fully optimized. Hopefully, this could be addressed with a firmware upgrade.—John Atkinson
Read more at https://www.stereophile.com/content...da-processor-measurements#jrceVBYdWVT2v1yd.99

There seems to be no love lost between JA and Schiit. I agree with JA's opinions about Ragnarok and how it works. I think Schiit should have done a more conventional design that doesn't behave "differently" when not listening to musical material. My amp shouldn't care if I'm listening to a sine wave or a frequency sweep. That's my choice.

 

[March Audio]

Or, you know, anyone who has to deal with the public. I have a rule; don't be the, "I want to speak to the manager," lady straight out of the gate. Whoever you are talking to is also a person and they are probably working in good faith. (At least, assume so at first.)

Really not sure I understand your response in this context. Who else is going to answer Jerrys questions other than Schitt support? Why shouldnt he contact them if it potentially doesnt technically perform to his expectation or their claims?

"Forget everything you know about DACs. Yggdrasil is the world’s only closed-form multibit DAC, delivering 21 bits of resolution"

As an aside, where I am, in Australia, consumer protection law is explicit and simple. I appreciate this isnt the same everywhere, but its really about companies conducting business in good faith.

ACCC
Replacements and refunds
You can ask for a replacement or refund if the problem with the product is major.

What is a major problem?
A product or good has a major problem when:

• it has a problem that would have stopped someone from buying it if they’d known about it
• it is unsafe
• it is significantly different from the sample or description
• it doesn’t do what the business said it would, or what you asked for and can’t easily be fixed.

 

[Wombat]

There seems to be no love lost between JA and Schiit. I agree with JA's opinions about Ragnarok and how it works. I think Schiit should have done a more conventional design that doesn't behave "differently" when not listening to musical material. My amp shouldn't care if I'm listening to a sine wave or a frequency sweep. That's my choice.

I am wondering why you are assuming a role here, which Schiit have decided they have no interest in pursuing. Unless you are a defacto insider you really have nothing to rationally defend or to add to the issue that hasn't been said here before. Your tenacity is misguided.

 

[garbulky]

I think the problem with your last sentence above is much of the gear has surpassed what humans can discern. Humans don't like being told that. So they search for what isn't there or bump into things that really do sound different because of an artful coloration (and therefore necessarily of lower fidelity).

Some fair amount of academic research into human hearing maps out at least in broad outline if not better things that can and can't be heard and at what levels. Audiophiles just will not be convinced by it vs subjective experience.

Without derailing the thread, let me quickly respond. We can take the rest up on PM if you like as it is somewhat off topic.

I'm not disputing that we can't hear 30 khz or that we can't hear 0.1% THD. I mean a way to quantify or describe our subjective experience of music itself - the more vague stuff.

For instance, how big is the soundstage. How forward in it is that violin going to play. How muddy does the orchestra sound at the back? Why does this sound clearer and in what way? As you can see these questions are impossible to answer with lots of specifics. We can say things like, the violin may sound overly bright due to a treble peak. Or it may sound more "dimensional" due to harmonics/"distortion" present in a tube amp.
They can go a step further and say if we add dsp effects to the signal we can make the sound seem more spacious. "Hall mode".

But further detailing is hard, at least in a clear standardised way that a standard person can reference similar to say a frequency response graph. It makes it harder that different people perceive things differently/pay attention to different stuff so a way to describe it represents challenges.

Right now the general idea is "you can't control how people perceive things, so why bother. Measurements don't change but people do. So focus on measurements." Which is understandable but if they could meld the two, flesh out more details, that would be great.

 

[Jimster480]

Check out baldr's posts on head-fi. He talks about the need for leaving it on. If you have issues with his explanation take it up with the designer. Oh and I agree with @amirm that it really needs to be mentioned in the manual and also preferrably on the web page.

I wouldn't say that this holds any weight. Devices warm up rather fast, especially without fans.
After a day it wouldn't matter either, devices wouldn't ever take so long to reach an optimal operating temperature.
I looked into the technical PDF's for those DAC chips they are using and nowhere does it mention needing to be running for any specific time to achieve performance. Just a range of operating temperatures like any other piece of equipment.

 

[Thomas savage]

Without derailing the thread, let me quickly respond. We can take the rest up on PM if you like as it is somewhat off topic.

I'm not disputing that we can't hear 30 khz or that we can't hear 0.1% THD. I mean a way to quantify or describe our subjective experience of music itself - the more vague stuff.

For instance, how big is the soundstage. How forward in it is that violin going to play. How muddy does the orchestra sound at the back? Why does this sound clearer and in what way? As you can see these questions are impossible to answer with lots of specifics. We can say things like, the violin may sound overly bright due to a treble peak. Or it may sound more "dimensional" due to harmonics/"distortion" present in a tube amp.
They can go a step further and say if we add dsp effects to the signal we can make the sound seem more spacious. "Hall mode".

But further detailing is hard, at least in a clear standardised way that a standard person can reference similar to say a frequency response graph. It makes it harder that different people perceive things differently/pay attention to different stuff so a way to describe it represents challenges.

Right now the general idea is "you can't control how people perceive things, so why bother. Measurements don't change but people do. So focus on measurements." Which is understandable but if they could meld the two, flesh out more details, that would be great.

You can create a thread if you like

 

[Touchstone]

One can lie with measurements easily by careful omission of facts or settings that are wrong for the type of measurement the instrument is doing, or with a poorly chosen instrument for the job. It's like lying with statistics.

Sorry, but unless you can demonstrate this to be the case concerning a particular (actual) measurement, then that measurement is a measurement and you've got no grounds for disputing it. Hypotheticals and conditionals and the use of the subjunctive mood aren't arguments against the accuracy of any measurement. Nor are similes. They're just rhetorical gestures.

 

[Thomas savage]

I am wondering why you are assuming a role here, which Schiit have decided they have no interest in pursuing. Unless you are a defacto insider you really have nothing to rationally defend or to add to the issue that hasn't been said here before. Your tenacity is misguided.

He’s just trying to make sense of the results here in the context of his own expirences and expectations. I think we should applaud those who come here, those with maybe different mind sets but want to consider what we are doing anyway..

They want to consider us the least we can do is consider them in return.

 

[Thomas savage]

Sorry, but unless you can demonstrate this to be the case concerning a particular (actual) measurement, then that measurement is a measurement and you've got no grounds for disputing it. Hypotheticals and conditionals and the use of the subjunctive mood aren't arguments against the accuracy of any measurement. Nor are similes. They're just rhetorical gestures.

He’s right in the respect someone with greater knowledge or who positions themselves as a ‘guru’ can use that position to manipulate others..

It’s perfectly possible to present data with a prejudicial slant in order to control people’s perception of the results.

It’s also perfectly possible to have two ‘ exsperts’ look at the same research and come to strikingly different conclusions leaving the ‘ non experts’ with the unenviable tasking choosing who to ‘ belive ‘ .

All theses kind of things are challenges we as a forum have to overcome but by far the best way is to educate ones self. If Audio is a hobby for someone why not geek out on the technical stuff, I’m told it can be fun

 

[GearMe]

...I'm not disputing that we can't hear 30 khz or that we can't hear 0.1% THD. I mean a way to quantify or describe our subjective experience of music itself - the more vague stuff...

..But further detailing is hard, at least in a clear standardised way that a standard person can reference similar to say a frequency response graph. It makes it harder that different people perceive things differently/pay attention to different stuff so a way to describe it represents challenges.

Right now the general idea is "you can't control how people perceive things, so why bother. Measurements don't change but people do. So focus on measurements." Which is understandable but if they could meld the two, flesh out more details, that would be great.

That's my thinking as well...just saying one set of measurements is significantly better is only part of the story. Being able to describe what the expected differences are in the listening results should be would be very useful (with allowances due to 'perception' differences as Garbulky mentions).

I was being dead serious when I asked the question regarding the THD+N differences [~50 dBs at 20 Hz!...Yggy (-61 dB@20 Hz), DX7 (-106dB@20 Hz) , E32 (-110 dB@20 Hz)]...which on the surface seems substantial! Surely, there must be some listening differences that align with those numbers when they're that disparate...yes?

 

[Wombat]

Thanks, Amirm. I have no dog in that fight...merely presented what Mike said. As stated, it's easy to show that the measurements haven't changed for the Yggy after 72 hours, and for the Mimby, it seems the thermal equilibrium time is a tad north of 2 hours...making the measurement task somewhat easier...all for the good...yes?

To my way of thinking, Mike 'throwing stuff out there' regarding a time-shift of 72 hours really doesn't buy him anything. Surely people are capable of leaving a device on for 72 hours prior to measuring it! In fact, if they measure it at 0 hours and then 72 hours...AND the measurements don't change, then Mike has the potential to look pretty foolish...especially if the tests can be replicated by others.

FWIW, the concept of an audio manufacturer having to back their 'folklore' by measurements and data is noble but not required...except by those individuals that deem it so for their buying criteria.

Engineering designers measure or they are just tinkerers.