• 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!

SMSL SU-8 Version 2 Balanced DAC

JohnYang1997

Master Contributor
Technical Expert
Audio Company
Joined
Dec 28, 2018
Messages
7,175
Likes
18,292
Location
China
To the people who think ess based dac measures good but sounds bad? Props for them for being honest and not blinded by objective environment. The reality is if they sound bad they must measure badly. Anything that's audible must be measurable. It's just often time we don't have the methodology to find the flaws of designs. When the design goes very far something can go horribly wrong without noticing. People aren't stupid.
 

JohnYang1997

Master Contributor
Technical Expert
Audio Company
Joined
Dec 28, 2018
Messages
7,175
Likes
18,292
Location
China
57f8628daddeb542.jpg
5803746abdc59833.jpg
6863e38d6cc5cdf8.png
-331b5fba2e8a1f83.png
1dc504b18b88055a.png
1dc504b18b88055a.png
-331b5fba2e8a1f83.png
-678b7014bf55e6c3.png
-6dbcdeb2bcd82af.jpg
-5845dea25a85a7a.jpg
28499d28fc4f9180.jpg
479ee53c54fc5440.jpg
4be925269ee706ba.jpg
 

JohnYang1997

Master Contributor
Technical Expert
Audio Company
Joined
Dec 28, 2018
Messages
7,175
Likes
18,292
Location
China
And there are some measurements that su8 looks good or ok.
-43a82811dabfcd62.png
-403b26f37c52c19a.jpg
261d44414c4b4e47.png
 

JohnYang1997

Master Contributor
Technical Expert
Audio Company
Joined
Dec 28, 2018
Messages
7,175
Likes
18,292
Location
China
Same test conditions as some of the measurements above but done with smsl m3.
280b51c98323aab4.jpg
24c95126ec2b52d4.jpg
35b4f5e4d706a1be.jpg
-50fcd02195abf9b6.png
 
Last edited:

JohnYang1997

Master Contributor
Technical Expert
Audio Company
Joined
Dec 28, 2018
Messages
7,175
Likes
18,292
Location
China
I also ran the same two tests with focusrite forte. Left is forte's input, right is using motu 828es input
-7a98fa0fcba4764c.png
-719428a613ec0113.png
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,394
Location
Seattle Area
To the people who think ess based dac measures good but sounds bad? Props for them for being honest and not blinded by objective environment.
Honesty has nothing to do with it. They need to run controlled listening tests before claiming such.

Your graphs have too little contrast with too small of fonts for me to read. Please pick one and explain what you are saying is audible in there and why.
 

JohnYang1997

Master Contributor
Technical Expert
Audio Company
Joined
Dec 28, 2018
Messages
7,175
Likes
18,292
Location
China
Honesty has nothing to do with it. They need to run controlled listening tests before claiming such.

Your graphs have too little contrast with too small of fonts for me to read. Please pick one and explain what you are saying is audible in there and why.
I don't want to talk about audibility here. But I'll explain what's happening.
Basically the main two tests that triggers weird patterns are:

Dual tone: 30hz + 1000hz with a ratio of 4:1

Multitone with pink tilt(like pink boise -20db/dec) that are NID in rew, or 1 5 12 29 49hz ... which is a irregular array of frequencies.

Lots of weird and unseen patterns appeared.

The graphs are just dual tone with different frequencies at different level, 1:4 or 1:1 etc. Just to make sure that this is not a special case in the dac.

And all the weird pattern does not appear on other dacs I have. Or at least magnitudes difference.
 

JohnYang1997

Master Contributor
Technical Expert
Audio Company
Joined
Dec 28, 2018
Messages
7,175
Likes
18,292
Location
China

Veri

Master Contributor
Joined
Feb 6, 2018
Messages
9,596
Likes
12,036
@JohnYang1997 while surprising the ESS chip seems to have such varying performance, much of what you measured seems to occur way too far down in level to be audible or to make one say "SEE, I knew ESS chips sound bad!". Especially the jitter you measured, seems just fine. No way it's audible.

I also think the ESS9028Pro/ESS9038Pro will measure better in these troublesome domains, the mobile chips can only go so far. For the average consumer I'm sure it's fine...
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,394
Location
Seattle Area
I don't want to talk about audibility here.
I thought you were providing these measurements to back what you said people say about ESS sound. I have not seen any measurements that back those statements in countless devices I have tested. Such people can't hear distortions far worse than these.
 

JohnYang1997

Master Contributor
Technical Expert
Audio Company
Joined
Dec 28, 2018
Messages
7,175
Likes
18,292
Location
China
@JohnYang1997 while surprising the ESS chip seems to have such varying performance, much of what you measured seems to occur way too far down in level to be audible or to make one say "SEE, I knew ESS chips sound bad!". Especially the jitter you measured, seems just fine. No way it's audible.

I also think the ESS9028Pro/ESS9038Pro will measure better in these troublesome domains, the mobile chips can only go so far. For the average consumer I'm sure it's fine...
I did blind test months back and the difference between various dacs are really easy to tell. Measurements are static signal measurements. You don't know what happens if a pulse injects. Also in a highly dynamic music, harmonic distortion is seems to be much more natural than all these weird stuff. In some cases, you can see the noise amplitude reference to the 1khz signal which can go up to -80db. Which is a wtf for almost state of art performance. In a dynamic situation, it's highly possible to be audible.
Anything wrong in the measurements will in some way reflect to the final sound. Don't just say it's inaudible. Even when it's inaudible, is there benefit to have such poor performance dac reference to cheap m3?
If anything under 0.1% thd is inaudible why would we want performance. These double standard needs to stop. People just want to stick to their original opinions. Bad is bad. Better is better. There are no excuses.
 

JohnYang1997

Master Contributor
Technical Expert
Audio Company
Joined
Dec 28, 2018
Messages
7,175
Likes
18,292
Location
China
I thought you were providing these measurements to back what you said people say about ESS sound. I have not seen any measurements that back those statements in countless devices I have tested. Such people can't hear distortions far worse than these.
I'll try to use real music for measurements. But it's really hard to analyze non periodic signal and highly dynamic signals.
 

JohnYang1997

Master Contributor
Technical Expert
Audio Company
Joined
Dec 28, 2018
Messages
7,175
Likes
18,292
Location
China
I thought you were providing these measurements to back what you said people say about ESS sound. I have not seen any measurements that back those statements in countless devices I have tested. Such people can't hear distortions far worse than these.
Blind test done. What a big difference.
Firstly the cymbals grabs my attention. The tone and texture are completely different on the two dacs su8 and m3.
Also the sound of other instruments from su8 kinda sounds soft/not solid as they should be and there is some "mp3" like sound to it.
However the first few seconds, su8 is very pleasing and has superb imaging. But then the coherence falls short.
There is no way anyone that can't tell these two apart. 5 seconds maximum. 100% accuracy.
 

Veri

Master Contributor
Joined
Feb 6, 2018
Messages
9,596
Likes
12,036
There is no way anyone that can't tell these two apart. 5 seconds maximum. 100% accuracy.

How are you sure volume is matched, though? Since this is pretty crucial in your blind test. (just honest question about your testing)
 

JohnYang1997

Master Contributor
Technical Expert
Audio Company
Joined
Dec 28, 2018
Messages
7,175
Likes
18,292
Location
China
How are you sure volume is matched, though? Since this is pretty crucial in your blind test. (just honest question about your testing)
Volume match for dac? no. I don't need to. I always turn the volume all the way down before switching.
 

Jaimo

Active Member
Joined
Feb 16, 2018
Messages
198
Likes
179
Location
Toronto, Canada

JohnYang1997

Master Contributor
Technical Expert
Audio Company
Joined
Dec 28, 2018
Messages
7,175
Likes
18,292
Location
China
So, you set the volume by ear?
Yes, basically. And I don't believe volume matching tests. The memory effect will ruin everything (even though maybe not this case).
The difference is much greater than the comparisons I did with khadas, m8a, mu2p.
To the point, I don't need to compare to tell which is which. I remember what each of the two dac sounds like.
 

Veri

Master Contributor
Joined
Feb 6, 2018
Messages
9,596
Likes
12,036
Volume match for dac? no. I don't need to. I always turn the volume all the way down before switching.

To the point, I don't need to compare to tell which is which. I remember what each of the two dac sounds like.

It is at this point, that we can figure... this is entirely your imagination. Volume match, I don't need to?
You can tell what each dac sounds like? What the hell man. I'm afraid you're overestimating yourself :oops:
 

JohnYang1997

Master Contributor
Technical Expert
Audio Company
Joined
Dec 28, 2018
Messages
7,175
Likes
18,292
Location
China
It is at this point, that we can figure... this is entirely your imagination. Volume match, I don't need to?
You can tell what each dac sounds like? What the hell man. I'm afraid you're overestimating yourself :oops:
No. easily. I wonder you guys actually listened?
 
Top Bottom