• 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 Soekris dac1421 Multibit DAC

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,835
Likes
16,498
Location
Monument, CO
1974, I think, and it looks like the "Text of Rule" link goes to my link. Great minds etc.? ;)
 

Jimster480

Major Contributor
Joined
Jan 26, 2018
Messages
2,880
Likes
2,033
Location
Tampa Bay
The line of thought is that the consumer who got a $1000 budget and is looking for an amp/DAC combo unit does not care about $2k+ standalone DACs, but instead if anything better than the dac1421 can be had for those money.
I get that they get compared in this review because they're both multibit, but from a consumer perspective, it's an apples and oranges comparison.
So then you want a DX7 since it spanks everything on this graph for half the price :)
 

soekris

Member
Joined
Feb 9, 2018
Messages
50
Likes
164
What filters does your DAC include? (Yes I could look it up, but...)

From the manual available on the soekris webste :)

“FILTER” LED color shows the current anti-aliasing filter selected
Red --- Linear Phase filter, also called brickwall filter
Orange --- Mix between Linear and Minimum Phase filter
Green --- Minimum Phase filter, also called butterworth filter
Off --- Soft Minimum Phase filter, a soft butterworth filter

Is there any advantage in your design to digitally limiting before the DAC core vs. just letting the DAC itself (R2R section) clip (limit)? I can think of pros and cons depending upon circuit implementation, just curious your thoughts on the subject.

Really appreciate the interaction and all of your time here!

Thanks,
Don

The limiter is saturating logic, have to be in the FPGA, logical place is just before going out to the R-2R resistors networks....
 

soekris

Member
Joined
Feb 9, 2018
Messages
50
Likes
164
So then you want a DX7 since it spanks everything on this graph for half the price :)

Some people can clearly hear a difference between DS DACs and R-2R DACs, but most can't can't hear any difference, they will just do fine with any crap coming out of China....

If my priority was the distortion specification I could just add a large amount of shaped noise and the FFT would be clean as the harmonic products would be spread out and moved up in frequency, just like a DS DAC do.... Yes, I have experimented with it, but see no improvements to the sound and a bunch of negative side effects....
 

andreasmaaan

Master Contributor
Forum Donor
Joined
Jun 19, 2018
Messages
6,652
Likes
9,399
Some people can clearly hear a difference between DS DACs and R-2R DACs, but most can't can't hear any difference, they will just do fine with any crap coming out of China....

If my priority was the distortion specification I could just add a large amount of shaped noise and the FFT would be clean as the harmonic products would be spread out and moved up in frequency, just like a DS DAC do.... Yes, I have experimented with it, but see no improvements to the sound and a bunch of negative side effects....

I've been interested in this question of why R2R allegedly sounds different from DS for some time.

In what measurement(s) does this audible difference show up in your opinion?

And what negative side effects did you encounter as a result of noise shaping?
 
Last edited:

Thomas savage

Grand Contributor
The Watchman
Forum Donor
Joined
Feb 24, 2016
Messages
10,260
Likes
16,298
Location
uk, taunton
I do hereby attest that work performed by The King Amir is worthy and all shall affirm that it is So.

Really, what Amir said... There are standards but many are out of date, such as the old stereo amplifier test criteria that do not encompass multi-channel testing or otherwise account for modern AVRs, there are no metrics I know for things like room correction, and no standards to say what must be reported beyond the bare bones. From an engineering standpoint datasheets for consumer products are very sparse in terms of real "meat" and things like THD vs. power specs are all over the map. This allows for market differentiation and gives the manufacturers leeway to report numbers they want to emphasize but make comparisons among products challenging.

Another example: There are some good data converter standards, like the IEEE ones I bring up now and again, but I do not think I have ever seen them mentioned in a consumer product test. They are required for some government programs, and referenced often in the academic and R&D world, but to the average consumer are way too deep and gov't regulatory bodies like the FTC do not enforce their use.

You need to understand enough technical detail to know when you can and can't compare things, and what allowances/changes to make to facilitate comparisons, but asking Amir to adhere to a standard where none exists is pointless (and at times degrading).

IMO! - Don
IMO too, thanks don.
 

soekris

Member
Joined
Feb 9, 2018
Messages
50
Likes
164
I've been interested in this question of why R2R allegedly sounds different from DS for some time.

I have a little page on my website: http://www.soekris.dk/tech_info.html

But it's a subject with many opinions....

In what measurement(s) does this audible difference show up in your opinion?

The current measurements are not able to show the shortcomings of a modern DS DAC as measurements are done using simple steady signals, typically one or two sinus tones, while music is much more complex and dynamic, and it is with complex signal that DS DACs fails as a high order DS modulator is basically unstable, imagine the digital tricks you need to do to get music out of a 1-5 bit DAC.... The first DS DACs was really crappy, some even had low level spurious tones, although they have come a long way since.

And what negative side effects did you encounter as a result of noise shaping?

Adding shaped noise (not the same as noise shaping)) did decrease the usable output level (and thereby decrease the S/N ratio) and added lots of high frequency noise, which can be filtered with additional output filters.... And then there is the added complexity and power use in the FPGA.
 

gvl

Major Contributor
Joined
Mar 16, 2018
Messages
3,427
Likes
3,982
Location
SoCal
Everyone who mentions that multibit sounds different than Delta-Sigma risks to be ridiculed here, but that's the opinion I formed after going through about 4 of each kind. YMMV.
 

dc655321

Major Contributor
Joined
Mar 4, 2018
Messages
1,597
Likes
2,235
The current measurements are not able to show the shortcomings of a modern DS DAC as measurements are done using simple steady signals, typically one or two sinus tones, while music is much more complex and dynamic, and it is with complex signal that DS DACs fails as a high order DS modulator is basically unstable, imagine the digital tricks you need to do to get music out of a 1-5 bit DAC....

Do you have suggestions for measurements that would demonstrate failures of DS DACs (or conversely, the superiority of R2R or multi-bit implementations)?

Here's one suggested several times on this forum, and elsewhere: record (digitize) the level-matched output from the devices under test and null against the original recording. The residuals are indicative of device distortion.

Music is just a signal - discrete representations of voltage. A DAC simply transforms samples into voltages. Signal complexity (deviation from pure tones) has nothing to do with that simple task.
 

andreasmaaan

Master Contributor
Forum Donor
Joined
Jun 19, 2018
Messages
6,652
Likes
9,399
The current measurements are not able to show the shortcomings of a modern DS DAC as measurements are done using simple steady signals, typically one or two sinus tones, while music is much more complex and dynamic, and it is with complex signal that DS DACs fails as a high order DS modulator is basically unstable, imagine the digital tricks you need to do to get music out of a 1-5 bit DAC.... The first DS DACs was really crappy, some even had low level spurious tones, although they have come a long way since.

Could you propose a hypothetical measurement, perhaps that doesn't use a sine signal as the input, that would demonstrate the claim?

Perhaps something involving an impulse, a traingular or square wave, etc etc?
 

garbulky

Major Contributor
Joined
Feb 14, 2018
Messages
1,510
Likes
827
@soekris
I too would be interested in a measurable way to show this difference. If you have any, please share. In my subjective (no DBT) experience I felt I kept hearing differences between the three multibit units I have had experience with and my DS DACs. But the measurements do not back me up So if you have some please feel free to share.
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,524
Likes
37,057
Do you have suggestions for measurements that would demonstrate failures of DS DACs (or conversely, the superiority of R2R or multi-bit implementations)?

Here's one suggested several times on this forum, and elsewhere: record (digitize) the level-matched output from the devices under test and null against the original recording. The residuals are indicative of device distortion.

Music is just a signal - discrete representations of voltage. A DAC simply transforms samples into voltages. Signal complexity (deviation from pure tones) has nothing to do with that simple task.

That is difficult to do. A good idea. The problem is timing. You have to compensate for clocks not being exact, and clocks of the ADC running at a different speed than the DAC. It is theoretically possible.
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,524
Likes
37,057
Apparently, 42 tone signals are inadequate as well, since my testing with them has failed to show up any of these hidden flaws.
Yes, those flaws, so incredibly evident upon hearing, are apparently well hidden beyond belief. As I've never seen anyone give the good measurements that show the superiority of multi-bit DACs. 1 tone is not enough or two. 6, 8 or even 42 not enough. But they must be keeping the measurements hidden alright.
 

pkane

Master Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
5,632
Likes
10,207
Location
North-East
That is difficult to do. A good idea. The problem is timing. You have to compensate for clocks not being exact, and clocks of the ADC running at a different speed than the DAC. It is theoretically possible.

Not just theoretically. A pretty good compensation is possible, as long as the relative clock drift is mostly linear.
 

gvl

Major Contributor
Joined
Mar 16, 2018
Messages
3,427
Likes
3,982
Location
SoCal
Given 2 DACs with I2S inputs it should be possible to clock them from outside using the same crystal, just need to make sure there is no additional ASRSC along the way.

Considering it is done, how does one interpret the delta?
 

gvl

Major Contributor
Joined
Mar 16, 2018
Messages
3,427
Likes
3,982
Location
SoCal
Subtract the two and then listen to the difference. Or analyze the difference with FFT and other tools to see what's in it.

Right, and what kind of conclusions can one draw from listening to or looking at the FFT of the difference? I don't think anyone expects the difference to be 0, which would be the only meaningful outcome.
 

March Audio

Master Contributor
Audio Company
Joined
Mar 1, 2016
Messages
6,378
Likes
9,317
Location
Albany Western Australia
@soekris
I too would be interested in a measurable way to show this difference. If you have any, please share. In my subjective (no DBT) experience I felt I kept hearing differences between the three multibit units I have had experience with and my DS DACs. But the measurements do not back me up So if you have some please feel free to share.
No DBT and "felt". No surprise then that the measurements didn't back up your biased and emotional response.
 
Top Bottom