This is a review and measurement of two versions of Yggdrasil: one with upgraded "Gen 5" USB board and the other with both Gen 5 USB and "Gen 2 Analog board." The latter is a bit of a misnomer because the board actually has the DACs on it which accept digital input. This is a "statement" DAC from Schiit and retail price of $2399 plus shipping. In the world of high-end that is a "bargain" and explains the popularity of this DAC.
Both of these units were made available to me at our local audiophile society meeting. I had to do my testing while other activity was going on so a bit stressful with some potential for error. After seeing some oddities I asked if I could remeasure both of them and both owners were kind enough to let me do that in a more relaxed setting. Each trip cost me 150 miles of driving for a total of 300 miles. Yes, you better feel sorry for me or else!
I know all of you are anxious about seeing the measurements of these two DAC variations. So here they are. Please read my tutorial on my setup and how these tests are conducted: https://www.audiosciencereview.com/forum/index.php?threads/understanding-audio-measurements.2351/
There is a lot of data here with two versions of these DAC. So I am going to show comparison to other DACs in a separate post although some of that is here too.
Measurements
Let's start with our favorite test of jitter and noise using 24-bit/48 Khz J-Test signal. You should only see a main spike at 12 Khz which represents our original tone. Here I am showing the difference between same test run over S/PDIF generated by my Audio Precision Analyzer and the Gen 5 upgraded-USB only Schiit Yggdrasil:
This is the only test where I saw a variation/improvement in the second test as shown on the graph. The low-frequency jitter reduced the second time around. Not sure what explains the variation. Regardless, the good news for users of this DAC is that its USB performance is now identical to S/PDIF. This is a first for Schiit DACs I have measured.
Alas we see some low frequency noise with either interface. I zoomed in to find the cause:
We see spikes at double the mains frequency of 60 Hz which indicates power supply noise bleeding into the output of the DAC. Due to lack of sensitivity in our hearing at low frequencies this is not much of an audible concern but should not be there to this level in such an expensive DAC. When we pay thousands of dollars for a DAC one of the expectations is power supply output so clean we could eat off of it. OK, that analogy doesn't work but you get what I am saying.
Comparing the Gen 1 analog board to Gen 2 in the second Schiit Yggdrasil we see an improvement:
Not textbook clean but good.
Let's look at linearity test or ability of the DAC to reproduce small input signals with fidelity:
Same theme emerges here with the Gen 2 board not having as wild of variation/error as Gen 2. That said, this is unacceptable resolution and fidelity for a DAC. I have tested much cheaper DACs that are almost ruler flat here, achieving nearly 20 bits of resolution. The ladder DAC used in Schiit DAC has an advertised performance of 20 bits so the implementation is way short of that. If you play high-resolution audio, this DAC is not for you.
Looking at linearity differently, this time seeing how it can reproduce the rightmost bits of a 16-bit audio sample (represented as 24 bits) at -90 dB or so, we get this:
Gen 2 analog board is definitely cleaner. It is less glitchy and noisy.
But wait, both of them have a problem in that the DC offset of the waveform is changing. We see that the waveforms are moving up and down. To understand this better, I ran the test for more cycles and got this:
Yup. The output is getting modulated by a sine wave around mains frequency! The modulation is significant relative to the amplitude of the waveform. Perhaps this is the bleed from power supply again. Either way, this is not acceptable in this price range DAC. It should have been instrumented and fixed.
Note again that both the new and old Analog boards have the same problem which makes sense as the power supply is unchanged.
Let's look at residual noise and distortion of a 1 KHz signal after the it has been filtered out:
The many variations in Gen 1 analog board shown in red is nicely reduced with the new Gen 2 analog board.
Let's now test the total amount distortion+noise relative to frequency:
Whoa! What went on here??? Your guess is as good as mine. The new board seems to be much worse the lower the frequencies get. And regardless, both are shown the door as compared to the Topping DX7 and Exasound E32.
Last but not least, let's take a look at frequency response measurement:
The top line is the Gen 2 analog board in balanced mode. All is well there. But if you look at the two curves at the bottom, both of which are for unbalanced output, we see problems. The Gen 1 board has a small roll off < 20 Hz which we could ignore. But Gen 2 board starts to drift down at some 300 Hz and by the time you get to 10 Hz, it is down by half a dB.
Controlled listening tests by Toole/Olive show that low frequency deviations that are this broad have a threshold of hearing of 0.5 dB. So this is right at threshold of hearing.
Note that these are stereo sweeps so both channels behave the same. This rules out a short, etc. in one of the channels. Whatever the cause, it is part of the output of two independent channels. Wonder if there is a manifestation of this in THD+N we saw earlier where Gen 2 unperformed Gen 1 analog board.
This was the main reason I wanted to remeasure the two Schiit Yggdrasil DACs. I couldn't believe a solid state DAC in this day and age can have such a frequency response error. Same results were confirmed in round two. I am at a loss as to how a DAC gets released which some frequency response error.
I should note that in the same sitting I measured the Berkeley Alpha DAC where we have ruler flat response. All that was done was I moved the cables and hit "go" again to remeasure. I also tested a Bryston BDA-2 with the same excellent outcome. So this rules out instrumentation error.
Conclusions
While the two versions of Schiit Yggdrasil produce the best performance I have seen from a Schiit DAC, they unfortunately point to performance problems that should not be there. Even if we put them aside, what I measured is simply not competitive. High-end audiophiles like to play high-resolution content and this is a DAC that struggles with 16-bit content let alone higher bit-depth.
I am sure we will be told that these two DACs are broken and that they need to go to service. Instead of creating work for me and their owners, manufacturer needs to replicate my tests and show that there are no problems.
For now, if you have done the Gen 2 analog board upgrade, I would contact Schiit and ask them whether they verified frequency response output before shipping it to you and if so, they can send you paperwork for that. This kind of thing better be part of the "pass/fail" testing of DACs before shipping.
And yes, I am aware of other measurements of Schiit Yggdrasil showing other performance. My answer to that is that Schiit is on record as saying that they make changes to their designs without notifying customers or changing model number/revision. So these errors could be recent. Don't quote me on this but I think the unit with just the Gen 5 USB is only 6 months old and the other one with analog upgrade ever newer as far as the DAC board is concerned.
Coming on the heels of the poor measurements of Schiit BiFrost Multibit DAC this is a much better offering. So if you are a fan of Schiit products and multibit technology, the Yggdrasil would be the way to go. But do not get the upgraded Gen 2 board or a new DAC with it without confirmation of frequency response error not being there. After all, that is an audible problem whereas the other improvements may very well not be.
Bottom line, I can NOT recommend the Schiit Yggdrasil DAC in old or new incarnations even though improvements are noted and appreciated over lower cost offerings from them.
As always, I welcome comments, corrections, questions, etc. I have been banned from both head-fi thread where the Schiit principles hang out so if you have a question or comment, please post it here. I have no ability to respond there.
My sincere thanks to the two local members of our audiophile society for letting me test their DACs not once but twice.
If you like this review, please consider donating funds for these types of hardware purchase using Patreon (https://www.patreon.com/audiosciencereview), or upgrading your membership here though Paypal (https://audiosciencereview.com/foru...eview-and-measurements.2164/page-3#post-59054).
Edit: fixed the comment about frequency response error.
Both of these units were made available to me at our local audiophile society meeting. I had to do my testing while other activity was going on so a bit stressful with some potential for error. After seeing some oddities I asked if I could remeasure both of them and both owners were kind enough to let me do that in a more relaxed setting. Each trip cost me 150 miles of driving for a total of 300 miles. Yes, you better feel sorry for me or else!
I know all of you are anxious about seeing the measurements of these two DAC variations. So here they are. Please read my tutorial on my setup and how these tests are conducted: https://www.audiosciencereview.com/forum/index.php?threads/understanding-audio-measurements.2351/
There is a lot of data here with two versions of these DAC. So I am going to show comparison to other DACs in a separate post although some of that is here too.
Measurements
Let's start with our favorite test of jitter and noise using 24-bit/48 Khz J-Test signal. You should only see a main spike at 12 Khz which represents our original tone. Here I am showing the difference between same test run over S/PDIF generated by my Audio Precision Analyzer and the Gen 5 upgraded-USB only Schiit Yggdrasil:
This is the only test where I saw a variation/improvement in the second test as shown on the graph. The low-frequency jitter reduced the second time around. Not sure what explains the variation. Regardless, the good news for users of this DAC is that its USB performance is now identical to S/PDIF. This is a first for Schiit DACs I have measured.
Alas we see some low frequency noise with either interface. I zoomed in to find the cause:
We see spikes at double the mains frequency of 60 Hz which indicates power supply noise bleeding into the output of the DAC. Due to lack of sensitivity in our hearing at low frequencies this is not much of an audible concern but should not be there to this level in such an expensive DAC. When we pay thousands of dollars for a DAC one of the expectations is power supply output so clean we could eat off of it. OK, that analogy doesn't work but you get what I am saying.
Comparing the Gen 1 analog board to Gen 2 in the second Schiit Yggdrasil we see an improvement:
Not textbook clean but good.
Let's look at linearity test or ability of the DAC to reproduce small input signals with fidelity:
Same theme emerges here with the Gen 2 board not having as wild of variation/error as Gen 2. That said, this is unacceptable resolution and fidelity for a DAC. I have tested much cheaper DACs that are almost ruler flat here, achieving nearly 20 bits of resolution. The ladder DAC used in Schiit DAC has an advertised performance of 20 bits so the implementation is way short of that. If you play high-resolution audio, this DAC is not for you.
Looking at linearity differently, this time seeing how it can reproduce the rightmost bits of a 16-bit audio sample (represented as 24 bits) at -90 dB or so, we get this:
Gen 2 analog board is definitely cleaner. It is less glitchy and noisy.
But wait, both of them have a problem in that the DC offset of the waveform is changing. We see that the waveforms are moving up and down. To understand this better, I ran the test for more cycles and got this:
Yup. The output is getting modulated by a sine wave around mains frequency! The modulation is significant relative to the amplitude of the waveform. Perhaps this is the bleed from power supply again. Either way, this is not acceptable in this price range DAC. It should have been instrumented and fixed.
Note again that both the new and old Analog boards have the same problem which makes sense as the power supply is unchanged.
Let's look at residual noise and distortion of a 1 KHz signal after the it has been filtered out:
The many variations in Gen 1 analog board shown in red is nicely reduced with the new Gen 2 analog board.
Let's now test the total amount distortion+noise relative to frequency:
Whoa! What went on here??? Your guess is as good as mine. The new board seems to be much worse the lower the frequencies get. And regardless, both are shown the door as compared to the Topping DX7 and Exasound E32.
Last but not least, let's take a look at frequency response measurement:
The top line is the Gen 2 analog board in balanced mode. All is well there. But if you look at the two curves at the bottom, both of which are for unbalanced output, we see problems. The Gen 1 board has a small roll off < 20 Hz which we could ignore. But Gen 2 board starts to drift down at some 300 Hz and by the time you get to 10 Hz, it is down by half a dB.
Controlled listening tests by Toole/Olive show that low frequency deviations that are this broad have a threshold of hearing of 0.5 dB. So this is right at threshold of hearing.
Note that these are stereo sweeps so both channels behave the same. This rules out a short, etc. in one of the channels. Whatever the cause, it is part of the output of two independent channels. Wonder if there is a manifestation of this in THD+N we saw earlier where Gen 2 unperformed Gen 1 analog board.
This was the main reason I wanted to remeasure the two Schiit Yggdrasil DACs. I couldn't believe a solid state DAC in this day and age can have such a frequency response error. Same results were confirmed in round two. I am at a loss as to how a DAC gets released which some frequency response error.
I should note that in the same sitting I measured the Berkeley Alpha DAC where we have ruler flat response. All that was done was I moved the cables and hit "go" again to remeasure. I also tested a Bryston BDA-2 with the same excellent outcome. So this rules out instrumentation error.
Conclusions
While the two versions of Schiit Yggdrasil produce the best performance I have seen from a Schiit DAC, they unfortunately point to performance problems that should not be there. Even if we put them aside, what I measured is simply not competitive. High-end audiophiles like to play high-resolution content and this is a DAC that struggles with 16-bit content let alone higher bit-depth.
I am sure we will be told that these two DACs are broken and that they need to go to service. Instead of creating work for me and their owners, manufacturer needs to replicate my tests and show that there are no problems.
For now, if you have done the Gen 2 analog board upgrade, I would contact Schiit and ask them whether they verified frequency response output before shipping it to you and if so, they can send you paperwork for that. This kind of thing better be part of the "pass/fail" testing of DACs before shipping.
And yes, I am aware of other measurements of Schiit Yggdrasil showing other performance. My answer to that is that Schiit is on record as saying that they make changes to their designs without notifying customers or changing model number/revision. So these errors could be recent. Don't quote me on this but I think the unit with just the Gen 5 USB is only 6 months old and the other one with analog upgrade ever newer as far as the DAC board is concerned.
Coming on the heels of the poor measurements of Schiit BiFrost Multibit DAC this is a much better offering. So if you are a fan of Schiit products and multibit technology, the Yggdrasil would be the way to go. But do not get the upgraded Gen 2 board or a new DAC with it without confirmation of frequency response error not being there. After all, that is an audible problem whereas the other improvements may very well not be.
Bottom line, I can NOT recommend the Schiit Yggdrasil DAC in old or new incarnations even though improvements are noted and appreciated over lower cost offerings from them.
As always, I welcome comments, corrections, questions, etc. I have been banned from both head-fi thread where the Schiit principles hang out so if you have a question or comment, please post it here. I have no ability to respond there.
My sincere thanks to the two local members of our audiophile society for letting me test their DACs not once but twice.
If you like this review, please consider donating funds for these types of hardware purchase using Patreon (https://www.patreon.com/audiosciencereview), or upgrading your membership here though Paypal (https://audiosciencereview.com/foru...eview-and-measurements.2164/page-3#post-59054).
Edit: fixed the comment about frequency response error.
Last edited:
