Schiit Yggdrasil -- 21 bit?


Schiit says that Yggdrasil is a 21 bit DAC. But the DAC chips that they put in the device ( Analog Devices AD5791BRUZ, 2 per channel) are 20 bit with the error of plus-minus 0.5 LSB.

How can the DAC be 21 bit if the chips are 20 bit? Using two chips per channel does reduce the RMS voltage of the noise by  a square root of 2. But how can you get to 21 bit from there?

Can someone please explain.
defiantboomerang
I believe Schiit addresses - at least somewhat - your question on their website FAQ page.

If you have a Yggdrasil, I'd be grateful for your opinion regarding my question(s) in this post https://forum.audiogon.com/discussions/new-yggdrasil-first-and-second-impressions
"How can the DAC be 21 bit if the chips are 20 bit?"

It can’t. John Atkinson review of the DAC clearly states that while the analog performance is excellent, the digital performance is less than desirable especially on high signal levels (which is the majority of pop rock music).

I wouldn’t want this DAC but I understand many here do not care about measured performance as long as they like the sound.
@shadorne 

I understand many here do not care about measured performance as long as they like the sound.

Yes, I am becoming aware of that. Could one say that such measurements-don't-matter audiophiles are like junk food addicts? They like to eat what tastes good to them without caring about the ingredients in their food?
@defiantboomerang 

Yes but I would say most audiophiles are dining on delicious gastronomic meals full of flavor and expensive ingredients rather than junk food.

I am interested in the accurate reproduction of the source file or music as produced by the artist and their engineers/producer but I fully appreciate why others like things to sound the way they want. After all at a restaurant you often order the dishes that you like best not necessarily what the chef has specially prepared.

@shadorne 

Good point. Those people do order expensive ingredients. Using their methods of analysis they can’t really tell the good ingredients from bad ones, though. Whatever sounds good is good. Whatever tastes good is good. 
@defiantboomerang / @shadorne 

Frankly, I'm interested, care about, etc. both measured performance and the sound. But as I've mentioned and am experiencing first hand (not to mention second hand by listening to other systems), I already have four DACs in total, but discounting the USB DAC in my M6si, that leaves two delta sigma (one in ERC-3 another in UDP-205) and the multibit in the Yggy.  They all sound great and none of them play a fourth verse in a song that only has three. So should I be led to believe that there is in fact a "nirvana dac" that measures perfectly and sounds perfectly so I can hear this fourth verse?
@gdhal 

Try the Benchmark DAC3. Nothing comes close. It will not give the fourth verse, but it will play your music better than anything else out there.

About all of your DACs sounding great -- of course they do. Nobody can tell decent modern DACs apart in a blind test. All perceived differences are just due to placebo.
"Nothing comes close."
Implicit: "...because, I’ve heard ALL the others and speak from first hand experience. "

"All perceived differences are just due to placebo."
Does that apply to the Benchmark DAC3 as well?
@defiantboomerang

I had given the Benchmark DAC 3 serious consideration. I talked at length with one of their sales rep, and I very much liked the fact that he spent the time with me. Something Schiit really doesn't do.

If I recall (and this was months ago so I'm going on memory only), for one thing the DAC 3 uses a Sabre chip 9028 whereas my Oppo uses the 9038. Granted, we all know there is more to the sound than *just* the chip (i.e. analog stage, etc.). Regardless, I just couldn't get past the fact that the Sabre chip in the Benchmark would sound like my Oppo, which uses an even more advanced chip.

Further, the DAC 3 has many "cousin" configurations in their lineup, neither of which zeroed in on the specific (and only specific) functions I was seeking - namely bit perfect PCM conversion. I couldn't care less about the headphone amp, home theater bypass and other features in the DAC 3 that I don't need/want. And because the Benchmark isn't really any less expensive than the Yggy,  it seems to me the cost is spread all that much thinner across all of its functions. So it's PCM specific abilities should be inferior to the Yggdrasil, because the Yggy's design focus is just PCM.

Unfortunately I had no way of listening to either one prior to purchasing, so the return policy of each was also something I considered.

Still, I think I made the wiser choice in my particular case to go with the Yggy.
@ghosthouse

"Nothing comes close."
Implicit: "...because, I’ve heard ALL the others and speak from first hand experience. "

No, nothing comes close to their technological expertise. DAC3 measures better than anything out there -- hence, nothing comes close.

"All perceived differences are just due to placebo."
Does that apply to the Benchmark DAC3 as well?

Of course it does. Tough, I will buy it, when I save some more money. Not because I think it will sound better than what I have, but because I like technically advanced devices. As regards its sound, the placebo effect for subjectivist audiophiles should/will come from the reviews that praise it. For example, the Stereophile says that DAC3 delivers "ASTONISHING FIDELITY AND EMOTIONAL EXPRESSIVENESS".
@gdhal

Yes, DAC3 and Oppo will sound the same. Not because they use similar chips, but because all decent modern DACs sound the same.

I am glad that you are happy with your purchase.

As regards bit perfect PCM -- the DACs in the Yggy use R2R ladders, at least that is my understanding. The precision of such DACs is not infinite or bit-perfect, because you cannot make perfectly precise resistors. To make a 20 bit DAC work, the resistors have to be really precise -- the tolerances should be less than one in a million and stable when temperature changes, that is < 0.0001%. Those are some accurate resistors! And even with very accurate resistors, the steps between bits will not be uniform, so there will be no bit-perfect output. I wrote a simulation program to see how the accuracy of resistors affects the output of such DACs -- with less accurate resistors you might not even get monotonic output. That is, you go one bit up digitally, but the output voltage goes down. Strange, but true.

Now, don’t get me wrong, the chips in the Yggy are excellent. But if you read their application note, you will see that they deliver about 19 bits in real life. And that is just the chips, the whole device cannot do better than that. DAC3 has just been reviewed by JA in the Stereophile. It delivers about 21 bits of resolution. Not that one could hear the difference between 19 and 21 bits.
Yes, DAC3 and Oppo will sound the same. Not because they use similar chips, but because all decent modern DACs sound the same.


@defiantboomerang

Interesting insight on your part that you bring to this thread discussion. So if the DAC3 and Oppo will sound the same, please allow me to ask a question that essentially goes back to square one. In your opinion, why should someone who currently owns an Oppo (and is happy with sound) purchase or even consider purchasing a DAC3?
@gdhal 

Because DAC3 is the pinnacle of technology at the moment. If it gives you satisfaction to use/own technologically advanced and superbly engineered devices, then DAC3 is your thing. If you don't care about specs and technology, then your choices might differ.
Post removed 
+1 to Dave and defiantboomerang seems more and more like a troll each passing day.
Well, I too would have to believe there is more to why someone would purchase a DAC 3 over a Oppo if sound were the primary consideration. 
Im using both the Yggdrasil and the Soekris dac1541 which is true R2R.
The Yggdrasil soundstage is phenomenal, i do however notice that mainly in electronic music the Soekris sound is more crispy and enjoyable. 



udishamir,
I would like to hear more about your impressions of the Soekris dac1541. Would you mind sharing?
Sure,

Its a very interesting DAC, its form factor is really appealing, its compact with an excellent sound quality. I must admit that before attaching it to the Lynx AES16e sound card (was using direct USB connection) it was OK, didn't "blow me away" (also leaving it working for few days does the trick). 

When connected with the Lynx sound card, then everything "opens up" started to hear new elements (especially in electronic music, soundtracks).

Soekris is a Danish company, i got the DAC two days after i ordered it (i live in California), They are very responsive.

My previous DAC was the Marantz HD-DAC1 which is a Delta Sigma, great sound for its price but beyond compare to the Soeakris or ofc the Yggdrasil which is a Monster, seriously. 

The Soekris is cheaper then the Yggdrasil and give it a fight :)

I personally recommend !

bellow is the dac1541 spec: 
http://www.soekris.dk/dac1541.html

Hope that helps.



It does. I researched them for a while but since it is new in the marketplace there were obviously no reviews. I ended up going with the Holo Audio Spring Dac but couldn't help but think for the money the Soekris represents a great value. Hope you enjoy it!
I do, love it.

While i tend to switch between the Soekris and the Yggdrasil constantly (depends on the style). Im waiting for the Spring DAC :)
Interesting in that Soekris states

 "It have the fantastic clean and natural sound of a 27 bit R-2R sign magnitude DAC..."

whereas Schitt states

"If your 24 bit recordings actually have 24 bits of resolution, we’ll eat a hat. And those "32-bit" DACs? Well, they have this measurement known as “equivalent number of bits.” This means, in English, how many bits of resolution they really have. And that number is 19.5. And 21 is better than 19.5, in all the math books we know."
Post removed 
16 bit for every problem,<100 dB resolution happy home.
17 kHz on max natural frequency in nature. The higher the resolution, the deeper the sound image of the natural sound section.

It can’t. John Atkinson review of the DAC clearly states that while the analog performance is excellent, the digital performance is less than desirable especially on high signal levels (which is the majority of pop rock music).

Hello @shadorne 

Please allow me to preface this post by stating I am *not* looking to be argumentative (with you are anyone else). And I do know you are aware that I have recently purchased the Yggy, and sincerely appreciate your feedback in another post of mine. :)

Your response to the OPs question "How can the DAC be 21 bit if the chips are 20 bit?" technically does not answer it. Irrespective of what JA claims, Schiit - the manufacturer of the product - clearly states "Yggdrasil is the world’s only closed-form multibit DAC, *delivering 21 bits of resolution* with no guessing anywhere in the digital or analog path. "

So, wouldn't you agree that at a minimum the Ops question should be posed to Schiit? Seems to me they - and not me, you or Atkinson - would be the best people to ask.
@gdhal

I don’t want to argue either. I posed a simple question in the OP. I got some good answers, but was also viciously attacked by some aggressive individuals here.

One’s logic almost compels one to ask: how can you get 21 bit resolution out of 20 bit DAC chips. Since independent verification is essential in science and engineering, JA’s (and other people’s) measurements are crucial. That is why we have reviews.

You are not against independent reviews, are you?
You can still get some signal resolution above 20 bits by dithering the 24 bit digital signal when converting to 20 bit. Dithering is much preferred to just truncating the least significant bits. Dithering adds random noise to the signal and raises the S/N but it will preserve some signal below the LSB. Studios do this all the time when taking 24 bit Masters and producing a 16 bit CD. 

I think Schiit are using the same logic as Meridan in saying that the last 3 bits on 24 bits is just noise (below current analog dynamic range resolution which is about 21 bit equivalent). This is true on a full scale signal. However a good recording may not use all full scale 24 bits so the last 4 bits could very well be within analog performance range and it is bad design to say we just throw them away. Also dithering is a form of processing and compression and just like with lossy audio and photo compression - some algorithms are bettter then others. 

Do do you trust your DAC manufacturer to process and massage the incoming 24 bit audio and throw away 4 bits or do you prefer to hear what the studio mastering engineer carefully produced?
How can the DAC be 21 bit if the chips are 20 bit?
The two DAC chips on each channel are described as being used in a balanced configuration. So I would suspect that one chip handles one half of the eventual output voltage range and the other chip handles the other half. Which would mean that 20 bits of resolution is provided for each half of the output voltage range, which in turn would correspond to 21 bits of resolution relative to the entire range.

That would be the theory, at least. How well it works out in practice will depend on the overall design, of course.

Regards,
-- Al

You are not against independent reviews, are you?

Of course not.

Your original question is valid, however, it hadn't really been answered (accurately or inaccurately) until after my post, when shadorne provided additional clarification and almarg offered yet another possible explanation.
@gdhal 

Sure thing. Well, I have learned a lot from some of the answers. I think that Schiit took a courageous step to use R2R chips in their DACs. They have one groovy machine there.
Putting one chip on -ve and one chip on +ve of balanced will simply double the voltage but won’t increase the bit depth. There really is no way to increase bit depth without a DAC chip with greater bit depth so I think Schiit is stretching the truth. I think that they are trying to say there is still signal below 20 bit due to dithering (random noise) that they must add prior to bit truncation of the last 4 bits of the 24 bit input signal. If they don't have the processing power to do dithering (normally done on a computer) then the audio quality will be compromised.
@shadorne 

Sorry, I am out of my depth here. Could you please help me with this. My thinking is as below. Where am I going wrong. 

Assume we have 21 bit of data. We use the MSB to select one of the DACs, the +ve or the -ve one. Then we send the remaining 20 bits to the selected DAC. Isn't that equivalent to sending 21 bits to a hypothetical 21 bit DAC? What am I missing?
I don’t think you’re missing anything, defiantboomerang.

Shadorne, if we denote the magnitude of the maximum possible output voltage as V, I’m envisioning that one DAC chip is controlling generation of output voltages between 0 and +V, with the other DAC chip generating 0 when a positive voltage is called for, while the other DAC is controlling generation of output voltages between 0 and -V, with the first DAC generating 0 when a negative voltage is called for. The outputs of the two DACs are then combined to create the overall output of the component. The voltage generated by each DAC would of course be quantized with 20 bit resolution, which would result in 21 bit resolution over the range from +V to -V.

Best regards,
-- Al

@almarg 

Not really. This just doubles the voltage value of each bit but it keeps the same resolution in terms as of number of bits or number of unique digital values available.
@defiantboomerang  

What you propose is indeed 21 bit. 20 bits from the two DACs each and an additional one bit MSB logic chip which controls whether the two DACs are additive or only one is turned on. This is not the same as balanced where both chips are operating simultaneously and are always on. 

I would call this a hybrid DAC - in fact using this topology you could take eight 16 bit DACs and couple them together with an 8 bit logic chip to create a 24 bit DAC.


Curious what, if anything, is wrong with my suggestion to pose the question to Schiit.
Hi Shadorne,

I'm not sure that what I said in my previous post came across clearly.  It amounts to the same thing the OP said, which you've agreed provides 21 bits of resolution, but described and perhaps implemented in a different manner.

What I described is not "balanced where both chips are operating simultaneously and are always on."  And it does in fact double the "number of unique digital values available."

Please re-read my previous post and see if you don't agree.

Best regards,
-- Al
 
@shadorne 

Let me ask you a different question. Is there a publication for rational audio hobbyists? I am getting tired of subjectivist nonsense. 

@almarg 

Yes. I did understand your post.

Glad we both agree that you can't magically get 21 bits from a 20 bit DAC without ADDITIONAL digital control (the extra bit) which then makes the DAC a 21 bit DAC not a 20 bit DAC. If you had added the extra MSB digital logic control explicitly then I would have understood beter what you were driving at. 






Really guys, 20bit-21bit.

Does it really matter, a PCM-1704 is 24bit, and your hard pressed to hear any difference to the almost swap-able (save for the package difference) 20bit PCM-1702 and that’s 4bit difference.

Cheers George
@georgehifi 

If playing CD quality files it really doesn't matter at all. Good point.

However in judging a product I do like to check manufacturers claims. For example technical claims by Hegel don't hold water. So either they got something messed up in translation from Norwegian or I have my doubts over technical competence.
I agree with @shadorne regarding (paraphrasing) drawing an inference to the product based on manufacturer technical claims. In fact, the boisterous claims by Schiit is one (of a few) reasons I purchased this particular DAC to begin with.

In addition to the boisterous claims by owners/reviewers :)
Post removed 
Quite awhile ago Jason explained all this in his "Schiit Happened"
Book he has on line at Head Fi. Too long to go through to provide an exact link. They sold a ton of these and are still selling well.
Don’t like don’t buy it!

You got Atkinson saying it's obsolete on the other hand you got Harley saying it's as good as a $19k DAC's!
Buffoons all of them and that's why I don't subscribe even for $7 a year . They can't give Stereopile away.