• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Review: Battle of Schiit Audio DACs

OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,386
Location
Seattle Area
Here is an interesting data point. I looked up the measurements of the Analog Devices AD5547 which Schiit Multibit uses: https://www.analog.com/media/en/technical-documentation/data-sheets/AD5547_5557.pdf

There, they show an FFT of THD similar to mine. Here is how their measurements compare to actual implementation in Schiit Multibit:

Schiit Modi 2 Multibit DAC AD Measurements.png


Notice the far, far better performance Analog Devices is showing in their measurements versus Modi Multibit. Their worst case distortion is around -110 dB or 25 dB better than Modi Multibit! That massive second harmonic peak seen in Modi Multibit is simply not there in Analog Device's measurements of the same chip in their bench implementation.

So we are really back to what I said in the review: people seem to sweep clear engineering problems away. They don't verify their designs properly and leave clearly engineering mistakes in there. Done correctly, the AD5547 would have been much closer to classic Delta-Sigma implementations. Until recently, Schiit did not take this seriously. They needed to have.
 

derp1n

Senior Member
Joined
May 28, 2018
Messages
479
Likes
629
Oh, yeah...duh :facepalm:

It looks like the MB has more ringing.
This is a result of the 18000 tap filter in their multibit products. You see a similar effect with Chord's bazillion-tap approach. A perfect filter would have infinite ringing, so this isn't a bad thing. If people really do hear differences in Schiit's multibit DACs, it's just as likely to be the filter than any "multibit magic".
 

Roen

Addicted to Fun and Learning
Joined
Oct 14, 2018
Messages
681
Likes
247
One correction,Modi 3 has SNR, not SINAD, of 111.
 

Rollaven

Member
Joined
Jun 5, 2018
Messages
39
Likes
19
@amirm,

Thanks for the schiit shoot out, love your site and the thorough reviews.

I had a question concerning the multiple tone test. So if I understood correctly if I would use an Topping DAC, or any other DAC with better measurements, instead of one of Schiit's I should here more detail in music? But what about the difference between D50 and D30 does one even hear these differences? And if so could you name an example of what kind of details one hears more or less with a better DAC in general?
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,386
Location
Seattle Area
That's what I figured, but is the total, the maximum output of the D/A or something else?
I did not look at the total output so don't know. Will look next time I am in the lab.
 

pwjazz

Addicted to Fun and Learning
Forum Donor
Joined
Sep 21, 2018
Messages
507
Likes
746
Anticipating the usual defense that sine waves are not music, I thought I do some further testing. Before that, remember, anything that distorts one tone, distorts all tones. In that sense, we are seeing actually less effect of distortion in above measurements than it exists with music. We can prove that by running our tests with multiple tones. Specifically, let's throw 32 tones at the DAC and see what happens (not all visible).

This is gold!
 

restorer-john

Grand Contributor
Joined
Mar 1, 2018
Messages
12,579
Likes
38,278
Location
Gold Coast, Queensland, Australia
Multitone!

My work here is done.

I want to get into Multitone too. I want to see real time effects of loading from volume pots in amplifiers/preamplifiers. Consider these DACs reviewed here are getting a nice 100K load from the AP when a lot of typical front ends of following stages are not that 'nice'.

What software should I be using to generate multitones? I have the complete Adobe CC but know nothing about Audition.
 
Last edited:

Jimster480

Major Contributor
Joined
Jan 26, 2018
Messages
2,880
Likes
2,032
Location
Tampa Bay
Awesome review!
Great to see the multitone tests to give people a better idea of how well it performs.
Your linearity test & Square waves do the same things though don't they? But I have argued this point here before about multi-tone tests, its great to see them used in this scenario. Before people start claiming multibit magic.
 

Jimster480

Major Contributor
Joined
Jan 26, 2018
Messages
2,880
Likes
2,032
Location
Tampa Bay
@amirm,

Thanks for the schiit shoot out, love your site and the thorough reviews.

I had a question concerning the multiple tone test. So if I understood correctly if I would use an Topping DAC, or any other DAC with better measurements, instead of one of Schiit's I should here more detail in music? But what about the difference between D50 and D30 does one even hear these differences? And if so could you name an example of what kind of details one hears more or less with a better DAC in general?
Yes you will, provided you have a good enough headphone & Amp. There are limits though, I know I could definitely tell the difference between the Modi Multibit vs my D30. I thought it weird that the Multibit sounded so bad when I read on Headfi that it was so magical.
Going back and forth I couldn't believe how bad it sounded, I lost so much low level detail in my songs that I had Schiit replace it with another one.
Only to have it perform identically to the second one.
 
Joined
Jun 19, 2018
Messages
56
Likes
89
Thanks as usual Amir!

So it seems there was actually a secret upgraded (fixed*) version of the Modi Multibit out there with no announcement or easy way to tell which version you have. Looks like SBAF got it right on that one. It seems your critique of the Yggdrasil trickled down into the Modi MB. I wonder if the MB Bifrost was the same? I would guess so based on the other MB products having the linearity issue.

I have heard people saying the MB provided more high end detail, so I assume they had it loud enough that the distortions became audible and made the sound brighter causing them to mistake increased distortion as high detail. Its nice to actually have a explanation for that.

I am curious, the 2nd harmonic distortion is pretty high in the MB, is this actually audible in any way? I tried one a few years ago and to me it sounded slightly bassier than my delta sigma dac, high 2nd harmonic distoriton would present itself that way if I am not mistaken.

I can also attest to the USB implementation of schiit dacs in general being bad, I have had issues with it not wanting to use wasapi and the asio driver they give sometimes gives me issues as well.

One of the big sells for their MB dacs is their comboburrito filter. does it alter the FR in anyway or is basically the same as a regular digital filter?
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,386
Location
Seattle Area
does it alter the FR in anyway or is basically the same as a regular digital filter?
Here is the frequency response difference between the two:

1543716470261.png


There is a 0.5 dB drop at 20 Khz which seems to be more than Modi 3 (red). This is from the multi-tone test so it may not 100% accurate. When I get a chance I will re-run using the traditional method.

The sample rate in this test by the way is 192 kHz.
 
Joined
Jun 19, 2018
Messages
56
Likes
89
Here is the frequency response difference between the two:

View attachment 18308

There is a 0.5 dB drop at 20 Khz which seems to be more than Modi 3 (red). This is from the multi-tone test so it may not 100% accurate. When I get a chance I will re-run using the traditional method.

The sample rate in this test by the way is 192 kHz.

Thanks for the response!

I think the comboburrito filter doesnt work at 192khz as I understand it. The DAC goes into some kind of nos mode at that point. It only works up to 96khz and below. WOuld be interesting to see how a sample rate below 192khz does.


Edit:

I was right about the Comboburrito, it doesnt work over 96khz, this is from the Bifrost MB but still applies

"For 176.4kHz and 192kHz input sample rates, Bifrost Multibit passes the music right along—no digital filter, no oversampling. Have fun! Upsample lower-rate music with another algorithm, and compare to our “comboburrito” filter."

I dont know if that impacts the MB performance at all, but would be interesting to test.
 
Top Bottom