• 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 and Measurements of Schiit Modi 3 DAC

Panelhead

Senior Member
Forum Donor
Joined
Apr 3, 2018
Messages
348
Likes
137
I have wondered if other frequencies for testing would expose more information.
I think the 1kHz was an early standard and it has been used since to allow apples to apples.
Some dacs really struggle with the 16 bit -90.3 dB sine wave. Every note played passes this level twice.
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,381
Location
Seattle Area
I ran into an article recently where one of the central findings was the idea that a high-end DAC had poor low-level linearity as compared to a cheaper, more linear one, but that dynamic compression could be perceived as a sonic benefit by many listeners.

https://hometheaterhifi.com/blogs/jj-s-now-and-then-blog-dac-linearity-and-perceived-audio-detail/
I read that but that is not anything reliable. Problem is in the comment section:

1539017357672.png


No way that kind of linearity is "hugely audible." He is performing subjective sighted listening tests and attributing what he thinks he has heard to a measurement. There is no cause and effect there.
 

mikehoopes

Member
Joined
Aug 2, 2018
Messages
82
Likes
94
Location
Campbell, CA
I read that but that is not anything reliable. Problem is in the comment section:

View attachment 16297

No way that kind of linearity is "hugely audible." He is performing subjective sighted listening tests and attributing what he thinks he has heard to a measurement. There is no cause and effect there.
He makes a claim there were no other measured differences. Perhaps they did sound different, but due to other, unmeasured factors (perhaps in other bands - maybe linearity was significantly worse above 1 KHz).
That's why I proposed characterizing the linearity artifacts, simulating them, and THEN listening to the differences. The effect could be varied until it was reliably identified in a DBT. Without a real control, confidence is low.
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,381
Location
Seattle Area
He makes a claim there were no other measured differences. Perhaps they did sound different, but due to other, unmeasured factors (perhaps in other bands - maybe linearity was significantly worse above 1 KHz).
I don't think they sounded different. Hearing differences at such low levels requires heroic skill and circumstances. No way can the outcome be as he described. He is just reporting faulty listening results.

That's why I proposed characterizing the linearity artifacts, simulating them, and THEN listening to the differences. The effect could be varied until it was reliably identified in a DBT. Without a real control, confidence is low.
That would be a good test to run! We may be able to simulate it with Matlab so can be done without hardware.
 

mikehoopes

Member
Joined
Aug 2, 2018
Messages
82
Likes
94
Location
Campbell, CA
I don't think they sounded different. Hearing differences at such low levels requires heroic skill and circumstances. No way can the outcome be as he described. He is just reporting faulty listening results.


That would be a good test to run! We may be able to simulate it with Matlab so can be done without hardware.
This makes me think of tone-mapping in HDR imaging processing to accommodate the difference between display contrast and our own internally post-processed visual luminance range; in the case of audio, you flip the equation. The reproduction equipment's capabilities may have outstripped our understanding of audibility, but solutions may be (or, perhaps unwittingly, have been) found to get those faint cymbal and foot-shuffling vibrations to pop, without otherwise painful side-effects.
[edit]: I'll take any further of my comments on this subject over to one of the DAC linearity threads, unless it deals specifically with the Modi 3. Thanks!
https://www.audiosciencereview.com/forum/index.php?search/34240/&q=linearity&c[title_only]=1&o=date
 
Last edited:

WetKneeHouston

New Member
Joined
Oct 10, 2018
Messages
1
Likes
0
Hey Amir, any chance you can measure jitter using optical? My motherboard's usb controller seems to hate the Modi 3 and coaxial isn't an option.
 

Veri

Master Contributor
Joined
Feb 6, 2018
Messages
9,596
Likes
12,036
Hey Amir, any chance you can measure jitter using optical? My motherboard's usb controller seems to hate the Modi 3 and coaxial isn't an option.
Since it's the same chipset doing the receiving/decoding work, just like with other devices it should be within inches of the coaxical performance.
 
Last edited:
Joined
Jun 19, 2018
Messages
56
Likes
89
Great review! Looks like your consumer advocacy might be having an effect on Schiit, good for everyone involved!

Schiit just updated their website with a APx555 Report for Modi Multibit , hope they do the same for all their products.

Made for an interesting read, its obviously worse than the Delta Sigma, but looks better than what some previous measurements showed.

Now SBAF has started a rumor that there is a Modi Multibit v2 which is presumably what those measurements are from, some of their testing suggests that Schiit updated the firmware at some point which improved its performance. Would be interesting if you could get a old Modi Multibit and test it versus a new one to see if there is any truth to this rumor.
 

SchwarzeWolke

Member
Forum Donor
Joined
Feb 7, 2018
Messages
94
Likes
82
And here we see why this forum or somebody like Nwavguy are so important. Shiit would have done his shit undetected for ages without those two!
Good to see that they are responding to real measurements. Pity that their attitude lives up to their company name... Just a pile of shit...
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,381
Location
Seattle Area
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,381
Location
Seattle Area
Hey Amir, any chance you can measure jitter using optical? My motherboard's usb controller seems to hate the Modi 3 and coaxial isn't an option.
Here you go:

1540004097398.png


It (in green) is actually better than both USB and S/PDIF! Likely the electrical isolation is helping it.

So go ahead and use it that way.
 
Last edited:

avatar

Member
Joined
Sep 2, 2018
Messages
13
Likes
12
Pretty impressive.
Why haven't you measured digital input on toslink instead of coax in all your past reviews ? (if available on the unit of course :D)
Could you measure sinad as well ?
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,381
Location
Seattle Area
Pretty impressive.
Why haven't you measured digital input on toslink instead of coax in all your past reviews ? (if available on the unit of course :D)
Could you measure sinad as well ?
Because USB is by far the most popular connection method. I only test other inputs by request.

On Schiit Modi, none of the interfaces are very clean so you see somewhat unique variations here.
 

villanova

New Member
Joined
Oct 20, 2018
Messages
1
Likes
1
Hi all,

Amir, thank you for your work. I created an account just to ask this question. Thanks to this review, I'm basically set on a Modi 3.

Are there any thoughts on Modi 3 versus Modi 3 Multibit? Is it just a better version of the 3?
 

RayDunzl

Grand Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
13,198
Likes
16,981
Location
Riverview FL
Because USB is by far the most popular connection method.

Gee...

TV, Oppo, HDRadio, Cable Box, DEQ2496, miniDSP OpenDRC-DI, CDPlayer - all output S/PDIF or TOSLINK

PC outputs USB

That's a popularity of 8:1 here against USB.

But I don't count, I know...

PS: The PC USB out gets converted to optical, so, make that 9:0 here.
 
  • Like
Reactions: pos
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,381
Location
Seattle Area
Are there any thoughts on Modi 3 versus Modi 3 Multibit? Is it just a better version of the 3?
I will measure the Multibit later but I suspect it will be awful compared to Modi 3. There is no reason whatsoever other than voodoo to buy Multibit. :)

Welcome aboard by the way.
 

mindbomb

Active Member
Joined
Nov 11, 2017
Messages
284
Likes
175
It (in green) is actually better than both USB and S/PDIF! Likely the electrical isolation is helping it.

So go ahead and use it that way.

That happened to some extent in the el dac review as well, that dac had a pulse transformer on the coaxial, and it too had lower noise on coax than usb. But the jitter was best with usb.
 

SchwarzeWolke

Member
Forum Donor
Joined
Feb 7, 2018
Messages
94
Likes
82
Funny, I'm sitting here in a seminar held by one of the developers from Steinberg in the 80-90s and who was responsible for the ASIO driver.
 
Top Bottom