Are Measurements of Schiit Yggdrasil DAC Inconsistent?

[andreasmaaan]

What I am saying is, parameters like dynamic range, distortion and S/N had hit the theoretical limits for 16/44 sources (for complete built up products). Now we have 24/192 sources and the D/A converters are nowhere near equivalent limits.

On a 16/44 source level playing field, modern D/As (including plenty of Amir's tests it would seem) are doing rather poorly in overall S/N. I've asked Amir to put 16/44 tests (as the majority of most people's sources are derived from 16/44 and that would allow meaningful comparisons with players from yesteryear) but he seems to like 24/44.

You cannot make a fair comparison between CD players D/As of the past and D/As of today without 16/44 tests, replete with 0dBFS THD and S/N.

Ok that makes more sense now.

Why do you believe many of the better-performing DACs Amir tests would not produce adequate measurements at 16/44?

 

[Blumlein 88]

Ooh I had a 500 and 563, my brother is still using them now. Wow, must be over 20 years old.

I used the 563 eventually with the odd Meridian 518. Among other things the attraction to me was it did digital volume control. I acquired an MSB ADC (an inexpensive one before MSB went crazy on pricing) thru which all my analog source went. Everything digital from there on out.

 

[Raimei Templar]

Appreciate your work here Amir, pretty well establishes that your measurements are consistent with others and dispels the critiques about your measurements being faked or wrong in some way.

Sorry you are taking a bunch of shit for it, but not surprised by it, there are a lot of exceedingly hostile people in this (and every) community.

It would be ideal if you could get the same units Atomic Bob and Jude measured as they could have indeed been given some sort of golden units that performed better, but clearly they arent willing to do that.

I bought and liked some of SChiits products before (Magni 3 and Modi MultiBit) but I have been very much turned off from them by their attitude towards you. I quite frankly cant understand why they are taking the position that their MB measure excellently.

Most people would be fine if they just said that Multibit dacs dont measure well but we think subjectively they sound better. That is a perfectly reasonable position to take. Instead their marketing is claiming 21 bits of resolution, which clearly it cannot do based on the measurements. Its being dishonest for no real reason.

 

[JP]

It’s pure marketing gold, though.

 

[restorer-john]

Why do you believe many of the better-performing DACs Amir tests would not produce adequate measurements at 16/44?

Take a sample of several modern D/A converter chips and peruse their respective datasheets. Look at the performance at 16/44, that is if they are brave enough to put the full specs for 16/44 in...

'Adequate measurements at 16/44' are really not what we are looking for are we?

 

[tranq]

New member, Schiit owner, Magni3, Modi Multibit. HD6xx. Question for @amirm.

Are the Schiit products measuring different then what they claim on their website, and what has been posted by the various reviews done on multiple sites? I've read most of your Schiit stuff, because Google News throws it in my face. The yggy gen 2 test you recently did seems to show Schiits measurements matching what they posted as far as specs go. I see you used a Topping DAC as a comparison.

Thanks.

I have been getting my feet back into audio the last half year, my older gear is 2ch and was purchased close to 20 years ago.

 

[amirm]

Are the Schiit products measuring different then what they claim on their website, and what has been posted by the various reviews done on multiple sites?

Welcome to the forum.

I have not done one for one analysis of their spec versus what I measure. Typical of many manufacturers, what they post is too terse, and too ambiguous to compare them to what I do. There is not one measurement graph of anything for any of their products.

But yes, the single figure THD+N did match my testing.

 

[tranq]

Welcome to the forum.

I have not done one for one analysis of their spec versus what I measure. Typical of many manufacturers, what they post is too terse, and too ambiguous to compare them to what I do. There is not one measurement graph of anything for any of their products.

But yes, the single figure THD+N did match my testing.

Thanks, I have a follow-up question(s).

It should be noted I have a beginners knowledge set in regards to audio, so sorry if these questions seem noobish.

How do your measurements relate to actual listening? I have seen a bunch of posts here related to single/double tone and quality of music, 44/96/192 etc.

Does the quality of the test tone have an effect on the data your analyzer kicks back?

Is the tone a file, or is it a CD? Or is the tone generated by the analyzer?

Are tones for testing standardized at all? Universal Review Standard?

If the tone is generated from the analyzer, would your tone be the same as one generated by a different reviewer with the same model analyzer?

Does anyone ever run a song through the DAC and analyzer to see how performance differs with music compared to test tones?

How does THD of a DAC, translate to music performance with an amp either HP or 2ch? My current 2ch amp has .2 percent THD. That tells me as a noob that I shouldn't worry so much about specs my ears can't hear, as my amps distortion will probably overlay any DAC distortion.

How does a DACs digital filter affect the test results it spits out from the analyzer? Again being a noob I would assume the digital filter is geared towards music and voice reproduction as opposed to tones reproduction.

Is there a link somewhere to an explanation of "everybody's favorite test"? Again noob card, and I've not read up on the linearity test.

Is it safe to assume that much of the differing opinions on here and head fi regarding Schiit (and other gear) is akin to the fanboydom gearheads are used to? An example would be proponents of your methods could be Porsche or BMW or Honda owners, and Schiit afficinados could be considered owners of Chrysler (scores poorly, unreliable, but people love their Jeeps), or Ford (reliable pickups and well made, but spartan interiors and overpriced) or Chevy (flashy and good looking, but not as reliable or well engineered as Honda), etc ?

I'm honestly trying to figure out my next purchase, (2ch) and I want to know how to fit your site into my purchasing process. I love the sound the Schiit stack gives to my music, I have added components in pieces and have heard a positive improvement with each piece. I put a good deal of time into reading reviews and commentary on each component before purchasing.

Thanks,

 

[derp1n]

Thanks, I have a follow-up question(s).

It should be noted I have a beginners knowledge set in regards to audio, so sorry if these questions seem noobish.

How do your measurements relate to actual listening? I have seen a bunch of posts here related to single/double tone and quality of music, 44/96/192 etc.

Does the quality of the test tone have an effect on the data your analyzer kicks back?

Is the tone a file, or is it a CD? Or is the tone generated by the analyzer?

Are tones for testing standardized at all? Universal Review Standard?

If the tone is generated from the analyzer, would your tone be the same as one generated by a different reviewer with the same model analyzer?

Does anyone ever run a song through the DAC and analyzer to see how performance differs with music compared to test tones?

How does THD of a DAC, translate to music performance with an amp either HP or 2ch? My current 2ch amp has .2 percent THD. That tells me as a noob that I shouldn't worry so much about specs my ears can't hear, as my amps distortion will probably overlay any DAC distortion.

How does a DACs digital filter affect the test results it spits out from the analyzer? Again being a noob I would assume the digital filter is geared towards music and voice reproduction as opposed to tones reproduction.

Is there a link somewhere to an explanation of "everybody's favorite test"? Again noob card, and I've not read up on the linearity test.

Is it safe to assume that much of the differing opinions on here and head fi regarding Schiit (and other gear) is akin to the fanboydom gearheads are used to? An example would be proponents of your methods could be Porsche or BMW or Honda owners, and Schiit afficinados could be considered owners of Chrysler (scores poorly, unreliable, but people love their Jeeps), or Ford (reliable pickups and well made, but spartan interiors and overpriced) or Chevy (flashy and good looking, but not as reliable or well engineered as Honda), etc ?

I'm honestly trying to figure out my next purchase, (2ch) and I want to know how to fit your site into my purchasing process. I love the sound the Schiit stack gives to my music, I have added components in pieces and have heard a positive improvement with each piece. I put a good deal of time into reading reviews and commentary on each component before purchasing.

Thanks,

This has all been covered before, on here and elsewhere. Please do some basic research.

 

[tranq]

This has all been covered before, on here and elsewhere. Please do some basic research.

I'm sorry, this floating header on the top of each page must have gotten me confused...

WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true).

 

[restorer-john]

Thanks, I have a follow-up question(s).

It should be noted I have a beginners knowledge set in regards to audio, so sorry if these questions seem noobish.

How do your measurements relate to actual listening?

Welcome to ASR.

Amir has a thread where the test conditions, equipment and tests performed are outlined. I can't seem to find it right now, but do a bit of a search.

As the site is primarily about actual measurements and the technical side of gear, the subjective side gets less attention. That said, at the end of the day, you can use the test reports, observations and opinions and then use your ears as the final arbiter on whether to lay down your cash or not.

At least that way, you go in with verifiable facts instead of just marketing words.

Come in and have a look around and ask any questions you like- we don't bite.

 

[TungstenC]

I'm sorry, this floating header on the top of each page must have gotten me confused...

Here’s the thread that you’re looking for https://www.audiosciencereview.com/forum/index.php?threads/understanding-audio-measurements.2351/

 

[SIY]

I love the sound the Schiit stack gives to my music, I have added components in pieces and have heard a positive improvement with each piece. I put a good deal of time into reading reviews and commentary on each component before purchasing.

There's an interesting truth in there that you may not have realized when you wrote this.

BTW, the car analogy fails when you compare relative prices.

 

[RayDunzl]

Does the quality of the test tone have an effect on the data your analyzer kicks back?

A "test tone" is often a sine wave, which can be created and the result examined with extremely high mathematical presicion. Garbage in, garbage out, as always.

Is the tone a file, or is it a CD? Or is the tone generated by the analyzer?

Depending on the input requirements of the gear under test, direct digital generation, digital files, CD with the file, or an analog tone could be required and usable.

If the tone is generated from the analyzer, would your tone be the same as one generated by a different reviewer with the same model analyzer?

A digital tone probably would, an analog tone would be subject to some variation.

Does anyone ever run a song through the DAC and analyzer to see how performance differs with music compared to test tones?

One of our members has performed DAC (Digital to analog) -> ADC (analog back to digital) in a loop of several generations before any audible difference was noted. The files are available on-site here someplace.

I love the sound the Schiit stack gives to my music, I have added components in pieces and have heard a positive improvement with each piece. I put a good deal of time into reading reviews and commentary on each component before purchasing.

Then you're good.

I'm happy with my setup, too. Worst rated speakers in a scientifical test. Antique preamp and amplifiers. Cheap as they come subs. Piss-ant DSP for room correction. Deaf ears, too.

 

[tranq]

There's an interesting truth in there that you may not have realized when you wrote this.

BTW, the car analogy fails when you compare relative prices.

I'm dumb, please elaborate.

 

[amirm]

How do your measurements relate to actual listening? I have seen a bunch of posts here related to single/double tone and quality of music, 44/96/192 etc.

The owner of the unit unfortunately wanted it back before I could finish my careful controlled testing. What I can say based on some amount of testing that neither the defects, nor the supposed advantages are audible for the music I tested. The Yggdrasil sounds similar to DACs at a fraction of its cost.

More careful testing and selection of more revealing material may have shown some differences. Whatever they are though, it is small.

 

[amirm]

Does the quality of the test tone have an effect on the data your analyzer kicks back?

No. The distortions in the analyzer itself are thousands of times lower than what is shown in Yggdrasil DAC.

Here is the performance of the generaator inside my analyzer and any distortion in the analyzer itself:

Notice how clean the spectrum of 1 kHz tone is in top right. Here is how the Schiit Yggdrasil did:

So the limitation is not remotely there in the analyzer.

 

[amirm]

Is the tone a file, or is it a CD? Or is the tone generated by the analyzer?

For one of the tests, Jitter, the data is in a file that is read and played by the DAC under test.

For other tests, digital samples are created on the fly and sent to the DAC being tested.

 

[amirm]

Are tones for testing standardized at all? Universal Review Standard?

Yes and no. There are common tests everyone runs such as THD+N. Or frequency response. For others, the parameters can vary. The purpose of this thread is that I showed even when parameters are varied, the outcome is always the same: the schiit Yggdrasil DAC does not perform well.

 

[amirm]

If the tone is generated from the analyzer, would your tone be the same as one generated by a different reviewer with the same model analyzer?

It would be the "same" because in the case of a DAC test, the samples are digital so not subject to variations in analog domain.

To be super technical, "dither" is used to generate proper samples. Dither is random noise so at the tiniest level, what is generated is a bit different from run to run. This variation is seen in the rough "grass" you see at the bottom of the graphs. We are not concerned about these variations though.