Budget DAC Review: Schiit Modi 2 ($99)

[Blumlein 88]

I know where you live.

You can't actually be sure though can you Sal. He has that nifty camper van and has the capability to go mobile. You'll never find him if he doesn't want you to find him.

That reminds me:

 

[Cosmik]

That reminds me:

Looking at that picture, it was a completely different time.

Forty years later we have Ed Sheeran.

 

[Sal1950]

Forty years later we have Ed Sheeran

Who?

 

[Nightlord]

Who?

 

[hvbias]

Thanks Amir.

Here are JA's measurements of the Yggydrasil, not particularly pretty either: https://www.stereophile.com/content/schiit-audio-yggdrasil-da-processor-measurements

And some measurements from someone on Headfi, though using less sophisticated tools than AP: https://www.head-fi.org/f/threads/yggdrasil-technical-measurements.764787/

Regarding this statement I've bolded and italicized from JA (on figure 6)

With undithered data and a signal at exactly –90.31dBFS, the Yggdrasil output a superbly symmetrical waveform, with the three DC voltage levels described by the data very well defined and the ringing due to the reconstruction filter clearly visible (fig.6). With undithered 24-bit data at the same level (fig.7), although the overall shape of the reconstructed sinewave is good, you can see significant errors at the signal's zero-crossing points. Again, this will be due to the design choice to use 20-bit converters.

Wouldn't the errors/glitching be more due to using non-audio DACs and a poorly done reconstruction filter?

There was a rather long discussion about this on Computer Audiophile and someone recorded some music out of the Yggydrasil and his PCM1704 DAC and I was able to tell them apart in a blind test, the Yggydrasil did not fare well.

 

[amirm]

Thanks for posting that. JA's review definitely shows anomalies with their design. Lack of monotonicity (increasing samples not resulting in increasing output)zero crossing error in an audio dac is bad form indeed.

BTW, the equipment used in the above screenshot is from Prism Sound and quite excellent.

 

[Blumlein 88]

There was a rather long discussion about this on Computer Audiophile and someone recorded some music out of the Yggydrasil and his PCM1704 DAC and I was able to tell them apart in a blind test, the Yggydrasil did not fare well.

I am no defender of the Schiit company. I would note however that file where you so clearly heard the Yggy different than another DAC, I heard it too, obvious in about 5 seconds. Well the reason is they used filtering on the other DAC that altered response in the upper mids and treble by lots. As in 13 decibels by 20khz. That is why you heard the difference. They had their preferred super musical (according to them) HQ player filtering rolling off the top end heavily. To post such a comparison was ridiculous. It wasn't the glitching as broken a design as I consider one with such glitches in this day and age that you heard.

 

[JoeWhip]

The guy at CA posting that nonsense had a clear agenda and thankfully finally stopped. Even the guy from Phasure had enough. I could say more but gave sworn I won't. It was nonsense. I will say that he complained how hard pianos sound through the Yggy. I was in a very fine recording studio at this time recording my son's last song, a piano/ vocal track on a just tuned Steinway. I played the piano a bit myself and was in the control room listening to the mic feed. We then listened to the file recorded at 24/96 WAV before sone small reverb was added. I took the files home and heard the same beautiful piano sound through my Yggy at home as I heard with the live mic feed. Yes, I have the raw and processed here at home no hardness or thinness at all. Diss Schiit all you want here guys, but the Yggy is a remarkable DAC. To hear the track for yourself albeit at CD Rez, go to John whip.com. The track is All Summer. Check out the other stuff too, including the single, Electric Love. Alive version is on YouTube also. No auto tuning!

 

[Blumlein 88]

The guy at CA posting that nonsense had a clear agenda and thankfully finally stopped. Even the guy from Phasure had enough. I could say more but gave sworn I won't. It was nonsense. I will say that he complained how hard pianos sound through the Yggy. I was in a very fine recording studio at this time recording my son's last song, a piano/ vocal track on a just tuned Steinway. I played the piano a bit myself and was in the control room listening to the mic feed. We then listened to the file recorded at 24/96 WAV before sone small reverb was added. I took the files home and heard the same beautiful piano sound through my Yggy at home as I heard with the live mic feed. Yes, I have the raw and processed here at home no hardness or thinness at all. Diss Schiit all you want here guys, but the Yggy is a remarkable DAC. To hear the track for yourself albeit at CD Rez, go to John whip.com. The track is All Summer. Check out the other stuff too, including the single, Electric Love. Alive version is on YouTube also. No auto tuning!

I don't know if the Yggy is a remarkable DAC. I doubt the glitches are audible. At the same time making something with them is completely un-necessary . So I am less than impressed by that design. The topic of this thread is another bit of Schiit gear that has some jitter problem. It may not be audible either, but again why build that as it is un-necessary to build a DAC with that problem these days. Plenty of good normal DACs are transparent to the source. What else should a good DAC do?

 

[DonH56]

I don't understand why the glitches would be due to using 20-bit converters instead of 24-bit, or whatever... I'm with Amir on that one; non-monotonic output is a horrible flaw and unacceptable for any DAC I've seen (audio or RF). Maybe at the lsb level, but would never be visible at that level.

 

[hvbias]

I am no defender of the Schiit company. I would note however that file where you so clearly heard the Yggy different than another DAC, I heard it too, obvious in about 5 seconds. Well the reason is they used filtering on the other DAC that altered response in the upper mids and treble by lots. As in 13 decibels by 20khz. That is why you heard the difference. They had their preferred super musical (according to them) HQ player filtering rolling off the top end heavily. To post such a comparison was ridiculous. It wasn't the glitching as broken a design as I consider one with such glitches in this day and age that you heard.

Woah I missed all that, at one point I decided not to read the thread due to all the fanboys from both sides. So were the Phasure files deliberately rolled off that steep to throw off the test or is that just the DACs normal response due to it being non oversampling? Was he exposed of this in that blind comparison thread or in another one?

IMHO the measurements of these DACs claiming all sorts of magic stuff is worth discussing and I personally wouldn't consider either the one Amir measured or the Yggydrasil high fidelity using its strictest definition.

 

[Blumlein 88]

Woah I missed all that, at one point I decided not to read the thread due to all the fanboys from both sides. So were the Phasure files deliberately rolled off that steep to throw off the test or is that just the DACs normal response due to it being non oversampling? Was he exposed of this in that blind comparison thread or in another one?

IMHO the measurements of these DACs claiming all sorts of magic stuff is worth discussing and I personally wouldn't consider either the one Amir measured or the Yggydrasil high fidelity using its strictest definition.

I'll pm you with some details when I return home.

 

[Cosmik]

I don't understand why the glitches would be due to using 20-bit converters instead of 24-bit, or whatever... I'm with Amir on that one; non-monotonic output is a horrible flaw and unacceptable for any DAC I've seen (audio or RF). Maybe at the lsb level, but would never be visible at that level.

I don't think it is necessarily non-monotonic; the glitch is a dynamic problem that occurs when many bits change simultaneously. If you buy an audio DAC IC for $1.50, the designers have anticipated that small problem, and fixed it for you.

 

[Cosmik]

With undithered data and a signal at exactly –90.31dBFS, the Yggdrasil output a superbly symmetrical waveform

I love the way they think this is some sort of achievement. The aforementioned $1.50 DAC IC will do that too.

Much 'high end' audio it seems to me is like this; people designing 'discrete' circuitry (cobbling together some obsolete industrial DACs, or building 'discrete' op amps, or relay-based attenuators) because it just has to better than those musicality-sapping integrated circuits and awful surface mount components. Then they measure their Heath Robinson contrivances using test gear built from ICs and surface mount components, and congratulate themselves if their circuit comes anywhere close to behaving like a $1.50 IC.

 

[Jinjuku]

Here's my question: With something like Oppo's recent $800 Sonica DAC, it's based on a cutting edge chipset, it's fully differential design and even the RCA outputs are derived from the balanced outputs, they've taken care of using a quality filtered, low noise power supply, etc... etc.... etc...

You can only engineer a circuit so far and then you are simply at the point of changing which chair you are sitting in on the Titanic.

What could a designer with a $5000 DAC possibly be doing that's actually of merit better?

 

[fas42]

You can only engineer a circuit so far and then you are simply at the point of changing which chair you are sitting in on the Titanic.

What could a designer with a $5000 DAC possibly be doing that's actually of merit better?

If you engineer a circuit so that it meets the requirements of a standard testing regime, nothing. If you want the component in the context of the system it forms part of, and in all the environments it finds itself, to correctly perform then a great deal more might be required.

IOW, should it be "just good enough" to keep the 'right' people happy, or should it be robust enough so that it always performs at optimum, no matter "what's thrown at it"?

 

[amirm]

Here's my question: With something like Oppo's recent $800 Sonica DAC, it's based on a cutting edge chipset, it's fully differential design and even the RCA outputs are derived from the balanced outputs, they've taken care of using a quality filtered, low noise power supply, etc... etc.... etc...

The oppo Blu-ray player that I tore down, see http://audiosciencereview.com/forum/index.php?threads/inside-of-a-premium-blu-ray-player.55/, had cheap Chinese electroylitic capacitors. So unless the DAC is different, they are cutting corners which will impact longevity.

 

[Sal1950]

What could a designer with a $5000 DAC possibly be doing that's actually of merit better?

A $4,500 custom machined aluminum case with a 3/4" thick magnesium faceplate?

 

[Sal1950]

["Right now, I swear, Schiit Audio's Mike Moffat and Jason Stoddard are sitting there in California, smugly smirking at me and John Atkinson. While JA was struggling to properly measure Schiit's Ragnarok (Fate of the Gods) integrated amplifier for my review in the May 2016 issue, I sent Moffat an e-mail: "Are you smiling?"

"Yup," he replied. He'd known in advance that the Ragnarok wouldn't look good on standard tests. But he hadn't warned us: The Ragnarok's output-stage bias program responds to music sources, not signal generators. "]

Is this a planned approach to further the audiohool agenda, pushing the position that science is useless in revealing what their magic dust products sound like?

 

[Blumlein 88]

The oppo Blu-ray player that I tore down, see http://audiosciencereview.com/forum/index.php?threads/inside-of-a-premium-blu-ray-player.55/, had cheap Chinese electroylitic capacitors. So unless the DAC is different, they are cutting corners which will impact longevity.

Having seen the inside of Oppo gear that is my opinion. I consider them a mass market company making products somewhere between good Sony and LG or Panasonic product. Somehow they got a rep started in high end audio they were doing more and just let it grow. So build quality is not always like other expensive gear. OTOH, if the sonics are good enough, and a digital product is likely obsolete for other reasons in 5 years and almost certainly obsolete after 10 years then more longevity just increases the up front cost to no real benefit.