Review and Measurements of Schiit Yggdrasil V2 DAC

[derp1n]

I really don't understand what (positive) thing Schiit is trying to show with that graph of atomicbob's in their marketing.

 

[Blumlein 88]

(Source)

While I now have quite a few albums with 24 bits, and more than 44.1 kHz, if I take into account my CDs from earlier days, and Spotify, the majority of the music I listen to is indeed 16 bit / 44.1 kHz. That's what the Yggdrasil is primarily designed for, rather than DSD, MQA or what have you.

Quoting from the same source as you.

Yggdrasil, or "Yggy," as the DAC in known informally at Schiit, supports all PCM formats from 16/44 to 24/192 through all inputs, including USB, AES/EBU, BNC, Coaxial, and Toslink optical. From there, the all-new USB Gen 3 input receiver and Schiit's proprietary Adapticlock™ system carefully manages clock regeneration for optimal performance. The SHARC DSP processor implements Schiit's proprietary 18,000+ tap digital filter algorithm at 352.8 or 384kHz sampling rate and 20-bit depth, which is then passed to the AD5791 DACs and a discrete JFET output buffer. Yggy outputs both balanced XLR and single-ended RCA (summed) analog simultaneously.

So they mean PCM, not DSD. They don't mean necessarily 16 bit 44.1khz. Though yes, that is also what most of my music is. I even agree if they properly arrived at good 20 bit performance is enough.

 

[Jacona]

I really don't understand what (positive) thing Schiit is trying to show with that graph of atomicbob's in their marketing.

A social experiment to see how many people are willing to just accept that something is good when they see a fancy graph that the average person would not be able to interpret.

 

[Blumlein 88]

I really don't understand what (positive) thing Schiit is trying to show with that graph of atomicbob's in their marketing.

They are trying to show a 1 khz tone at about -144 db implying 24 bit performance.

 

[Ron Texas]

I want 21 bits. Don't know what I will do with it...

 

[rebbiputzmaker]

Those weren't my measurements.

Then is it from external causes/environment?

 

[lotse888]

Then is it from external causes/environment?

I thought you suggested that it was an "internal issue" ==>"Are internal components/design causing the power supply noise you are measuring? "

 

[amirm]

Then is it from external causes/environment?

What?

 

[gvl]

They are trying to show a 1 khz tone at about -144 db implying 24 bit performance.

Ignoring the noise for the moment, how is it even possible with 20 bit DACs? Are they really showing just the LSB-level signal that registers at -144db level due to non-linearity?

 

[amirm]

I really don't understand what (positive) thing Schiit is trying to show with that graph of atomicbob's in their marketing.

It is engineers trying to be marketing people. As you say, no one has any idea what that graph is. And the people who do, know that it is mostly bad.

One thing is clear though: no one better talk about how this DAC "sounds." The company clearly wants their marketing message to be measurements and exactly the type we perform here! Actually, they are worse going to -144 dB and such, where any relationship with what we hear is long, long lost.

 

[amirm]

Ignoring the noise for the moment, how is it even possible with 20 bit DACs? Are they really showing just the LSB-level signal that registers at -144db level due to non-linearity?

For the same reason this "1-bit" image conveys shades of gray:

The key is dither. We add noise and extract more resolution.

 

[rebbiputzmaker]

What?

Those were both questions. "?" Sorry if they were not clear enough.

I would guess it is one or the other. (hopefully not a poltergeist) internal or external.

Wouldn't some further research be useful? i.e. clean isolated power to see if anything changes.

 

[amirm]

Wouldn't some further research be useful? i.e. clean isolated power to see if anything changes.

I have measured these devices three times, in three different environments/powers. Same issues remain.

If isolation power helps, manufacturer should experiment and let their users know. My job here is not to troubleshoot their problems. That is their job to do.

 

[rebbiputzmaker]

I have measured these devices three times, in three different environments/powers. Same issues remain.

If isolation power helps, manufacturer should experiment and let their users know. My job here is not to troubleshoot their problems. That is their job to do.

Using isolation transformers is common practice for sensitive laboratory or medical equipment for example. Any differences that may be observed, would be educational. I do not believe any manufactures asked you to troubleshoot their problems, just thought your efforts were an educational exercise.

 

[Blumlein 88]

Ignoring the noise for the moment, how is it even possible with 20 bit DACs? Are they really showing just the LSB-level signal that registers at -144db level due to non-linearity?

Amir has already answered. It is dither. It's possible to encode a -120 db tone in dithered 16 bit and recover it. Without dither it disappears as you'd expect. I've posted exactly that here. Maybe this will make it clear.

https://audiosciencereview.com/forum/index.php?threads/philips-red-book.2138/post-57952

And here is a post with zip files that unzip so you can listen to a -118db tone for yourself. I applied 60 db of digital gain so you can easily hear the tone is there among the dither. One dither is more effective than the other. You can hear this also.

https://audiosciencereview.com/forum/index.php?threads/philips-red-book.2138/post-57959

bennetng used a more aggressive noise shaped dither to get a -125 db tone recoverable.

https://audiosciencereview.com/forum/index.php?threads/philips-red-book.2138/post-57963

 

[amirm]

Using isolation transformers is common practice for sensitive laboratory or medical equipment for example.

It is? My wife worked in the hospital laboratory and they didn't use isolation transformers. Nor have I seen isolation devices while in hospital rooms. You have some experience you want to share on that ground?

Any differences that may be observed, would be educational. I do not believe any manufactures asked you to troubleshoot their problems, just thought your efforts were an educational exercise.

And I can educate by testing the numerous other gear waiting for measurement.

 

[Blumlein 88]

I've spent lots of time in hospitals and medical labs recently. I've seen things like this.

https://www.amazon.com/Tripp-Lite-IS1800HG-Isolation-Transformer/dp/B00008N6S7

Often tucked away if you aren't looking carefully. Usually in ICU's, not usually in regular hospital rooms.

I've also worked in non-medical labs and at times been the person ordering the gear. Some isolation gear was used for some purposes. Mostly to protect the lab equipment more so than to improve its performance.

This brand is one I've seen a number of times.
http://www.controlledpwr.com/SureImage_Medical_Power_Conditioner_Ultra-KM

These usually combine conditioning, isolation and UPS capabilities. If you look around most of your imaging gear, you'll see something like this. It may be in a closet, or behind a wall, but rarely (perhaps never) are such things as MRI, PET and CAT scan gear simply plugged into or wired into the wall. Again I think this is mostly to protect such gear more so than to improve its performance. I also imagine a hospital building full of the electronics you see in modern times is a far noisier environment than most people's homes.

 

[gvl]

Amir has already answered. It is dither. It's possible to encode a -120 db tone in dithered 16 bit and recover it. Without dither it disappears as you'd expect. I've posted exactly that here. Maybe this will make it clear.

https://audiosciencereview.com/forum/index.php?threads/philips-red-book.2138/post-57952

And here is a post with zip files that unzip so you can listen to a -118db tone for yourself. I applied 60 db of digital gain so you can easily hear the tone is there among the dither. One dither is more effective than the other. You can hear this also.

https://audiosciencereview.com/forum/index.php?threads/philips-red-book.2138/post-57959

bennetng used a more aggressive noise shaped dither to get a -125 db tone recoverable.

https://audiosciencereview.com/forum/index.php?threads/philips-red-book.2138/post-57963

Thanks, got it. I knew and it kind of made sense dither was needed when reducing the bit depth to minimize quantization error. Looking from the FFT perspective makes it more clear why it is necessary.

 

[gvl]

I've spent lots of time in hospitals and medical labs recently. I've seen things like this.

https://www.amazon.com/Tripp-Lite-IS1800HG-Isolation-Transformer/dp/B00008N6S7

Often tucked away if you aren't looking carefully. Usually in ICU's, not usually in regular hospital rooms.

I've also worked in non-medical labs and at times been the person ordering the gear. Some isolation gear was used for some purposes. Mostly to protect the lab equipment more so than to improve its performance.

This brand is one I've seen a number of times.
http://www.controlledpwr.com/SureImage_Medical_Power_Conditioner_Ultra-KM

These usually combine conditioning, isolation and UPS capabilities. If you look around most of your imaging gear, you'll see something like this. It may be in a closet, or behind a wall, but rarely (perhaps never) are such things as MRI, PET and CAT scan gear simply plugged into or wired into the wall. Again I think this is mostly to protect such gear more so than to improve its performance. I also imagine a hospital building full of the electronics you see in modern times is a far noisier environment than most people's homes.

People's homes have dimmers and LED lighting these days, they are inherently noisy.

 

[Blumlein 88]

People's homes have dimmers and LED lighting these days, they are inherently noisy.

I don't think it was here on ASR. But when I moved a couple years back I measured some noise basics before and after swapping all the bulbs from incandescents to LED. No change. I also measured the noise on the outputs of some audio gear and again no change. I don't have any dimmers so maybe that happens. Still I don't believe the grid in homes will approach that of a hospital with thousands of electronic devices plugged in.

Now LEDs can be electrically noisy of course. It just seems that noise doesn't travel that far or infect audio gear as often as expected. I do have a couple LEDs that create an audible high pitched buzz. And once had a triple bulb fixture with two different brands in it. All three buzzed at about 11 khz faintly. I changed the bulb out so all were the same brand and it went away. The removed bulb by itself in other location was quiet. So such things are more possible than in the past.