• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

Measurement of R2R DAC

DonH56

Technical Expert
Technical Expert
Patreon Donor
Joined
Mar 15, 2016
Messages
3,166
Likes
3,866
Location
Monument, CO
10.5 bits is 1 part in 1448 or 0.07% so in terms of something like THD you could argue it is below our threshold detection for a full-scale signal. But it also means noise or whatever is only 63 dB down so may be audible at lower levels depending upon the cause of the distortion and how it tracks with amplitude. The point I (and many others) have made before is that either way, given the number of inexpensive DACs that measure much (MUCH!) better, why pay for one that does not?

I suspect "Amir bits" is in reference to Amir's linearity deviation plots which are not based on ENOB per-se (he does not measure ENOB directly IIRC and the metric for deviation is not 6 dB but something lower). I have known and used ENOB for ages but it is not the end-all be-all spec and as long as I know what Amir is doing his scheme provides a good relative reference for comparisons among products. Which is all most of us want anyway.

IMO! - Don (well, the math is not opinion, but check it...)
 

derp1n

Senior Member
Joined
May 28, 2018
Messages
481
Likes
584

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
21,297
Likes
28,766
Location
Seattle Area
Reading the history of that product, seems like he farmed out the design to someone else anyway. And that person peaked at the design of the other product they say was copied. So he really doesn't know the signal processing or the design of the unit. The measurements I see elsewhere look pretty bad too. Shame because I was tempted to buy it to test.
 

esm

Active Member
Patreon Donor
Joined
Apr 9, 2018
Messages
118
Likes
149
Location
Berkeley, CA USA
Interestingly (to me, anyway) is that the original drop is targeting 960 units, presumably part of a strategy of paying for initial tooling, a production run, additional inventory, etc.

With 10 days left on the drop, they've only sold 236 so far. I assume there'll be a rush of folks clicking the button as it gets closer to finishing, but still, that seems pretty low to me.
 

AntonK

New Member
Joined
Jul 24, 2018
Messages
3
Likes
1
I heard an opinion that there is no pre-ringing in NOS DACs. And that for our ears / brain the lack of pre-ringing is more important than 10x higher distortion figures.
Can it be true?
Also, the distortion of nos DACs lies mostly in the higher frequencies (correct me if I am wrong). This is where our hearing is least sensitive.
So can it be that distortion is overrated as a sq criteria?

IMHO it is always important to discuss measurements in connection with psychoacoustics.

I used to have a cd which had the same track recorded with added even harmonics. 0.01%, 0.1%, 1%, 10%.
The purpose of the cd was to demonstrate how our hearing system works.

I could only reliably identify a track with added 10% harmonic distortion!
All other tracks sounded the same to me.

I own a nos and a well measured delta-sigma dac. The nos dac (metrum octave) is definitely more pleasant to listen to, especially in the midrange.
 

derp1n

Senior Member
Joined
May 28, 2018
Messages
481
Likes
584
I heard an opinion that there is no pre-ringing in NOS DACs. And that for our ears / brain the lack of pre-ringing is more important than 10x higher distortion figures.
Can it be true?
Also, the distortion of nos DACs lies mostly in the higher frequencies (correct me if I am wrong). This is where our hearing is least sensitive.
So can it be that distortion is overrated as a sq criteria?

IMHO it is always important to discuss measurements in connection with psychoacoustics.

I used to have a cd which had the same track recorded with added even harmonics. 0.01%, 0.1%, 1%, 10%.
The purpose of the cd was to demonstrate how our hearing system works.

I could only reliably identify a track with added 10% harmonic distortion!
All other tracks sounded the same to me.

I own a nos and a well measured delta-sigma dac. The nos dac (metrum octave) is definitely more pleasant to listen to, especially in the midrange.
http://archimago.blogspot.com/2018/01/audiophile-myth-260-detestable-digital.html
 

gvl

Major Contributor
Joined
Mar 16, 2018
Messages
1,289
Likes
858
Location
SoCal
The main trouble with NOS imo is IMD of image frequencies that may fold into audible band. Well, and the HF droop
 

andreasmaaan

Major Contributor
Joined
Jun 19, 2018
Messages
3,517
Likes
2,767
Delta-sigma designs tend to handle impulses more poorly than more conventional architectures so are not well-suited for some applications (or require special design techniques). For audio there are noise modulation and transient issues that delta-sigma designs are more susceptible to, but there are modulation-like and transient issues conventional designs also exhibit, so I am not sure there is a clear winner.
Sorry to revive an old post Don. I'm just wondering, what measurements would show up the noise modulation and transient issues that DS DACs are more susceptible to? And are these issues that you would expect to cause audible problems in well-designed DS DACs?
 

DonH56

Technical Expert
Technical Expert
Patreon Donor
Joined
Mar 15, 2016
Messages
3,166
Likes
3,866
Location
Monument, CO
Sorry to revive an old post Don. I'm just wondering, what measurements would show up the noise modulation and transient issues that DS DACs are more susceptible to? And are these issues that you would expect to cause audible problems in well-designed DS DACs?
How old is that? :)

A few years ago my job was designing data converters, both conventional and delta-sigma. RF delta-sigma designs are an interesting challenge. That's a polite way of saying they are a PITA with a number of quasi-hidden "gotcha's" it took a lot of work to ferret out.

Most any signal will show the noise modulation but DC or very LF signals tend to excite some of the filter modes and multitone or NPR (noise power ratio) tests will also show it (though may take a long'ish record). Impulse or step response tests, or square waves with fast edges, will show transient issues. Testing for things like this can get complicated and the causes of problems can vary (in any design, delta-sigma or not).

Delta-sigma designs are pretty ubiquitous these days and most of the early issues have been solved through features like noise decorrelation (dither), improved multi-bit loop architectures and better system analysis tools, better digital filters (and more powerful processors to run them), etc.

I would not expect most of these to be audible in most DACs but designs and opinions vary.

It is also worth noting that similar issues can arise with conventional designs. And often they have similar causes and solutions.

HTH - Don
 

andreasmaaan

Major Contributor
Joined
Jun 19, 2018
Messages
3,517
Likes
2,767
@amirm have you considered adding some kind of transient-based test to your repertoire? This seems to be the area in which DS DACs are most criticised by the R2R crowd, and @DonH56 seems to think there might be some basis for that. Having some data on this might be interesting and useful...
 

gvl

Major Contributor
Joined
Mar 16, 2018
Messages
1,289
Likes
858
Location
SoCal
Impulse response depends on the type of digital and maybe analog filter, if any, and not necessarily on the DAC type. All oversampling DACs will show some type of ringing with less than perfect impulse response and NOS R2R DACs will show none or minimal ringing and near perfect impulse response. This behavior is expected and well understood. While an interesting metric that can provide some insight on the implementation I don't see if there's a good way to draw conclusions on the merits of a particular device through comparing their impulse responses.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
21,297
Likes
28,766
Location
Seattle Area
While an interesting metric that can provide some insight on the implementation I don't see if there's a good way to draw conclusions on the merits of a particular device through comparing their impulse responses.
That's my opinion too. I swear reviewers plot those impulses and square waves to impress the readers that they know something the reader doesn't know! :)
 

andreasmaaan

Major Contributor
Joined
Jun 19, 2018
Messages
3,517
Likes
2,767
Impulse response depends on the type of digital and maybe analog filter, if any, and not necessarily on the DAC type. All oversampling DACs will show some type of ringing with less than perfect impulse response and NOS R2R DACs will show none or minimal ringing and near perfect impulse response. This behavior is expected and well understood. While an interesting metric that can provide some insight on the implementation I don't see if there's a good way to draw conclusions on the merits of a particular device through comparing their impulse responses.
Interesting. Regarding NOS R2Rs, it seems to make sense to me why the ringing would be mostly absent, but I would have thought this would be a result of the lack of a LPF on the output. Is it actually something inherent to the R2R design?

And are we talking about ringing in the audible spectrum (say under 20KHz) with DS DACs, or are we simply talking about ringing in the transition band of the LPF? And where a LPF is used in an R2R DAC, is there any difference?

Finally does anyone know of any measured data to demonstrate these differences?

I keep thinking I'm close to understanding the time domain effects of various DAC implementations, only to realise I'd misunderstood... :confused:
 

gvl

Major Contributor
Joined
Mar 16, 2018
Messages
1,289
Likes
858
Location
SoCal
Correct on LPF, but even with analog LPF there will only be post-ringing which is said to be less noticeable as it is masked by the signal. Recent NOS R2R designs avoid the LPF altogether for this same reason and also to not change frequency/phase response, or use a very gentle LPF filter, so ringing is minimal. Technically ringing is ultrasonic so should not be audible, but dome disagree. There was a relevant discussion here:
https://audiosciencereview.com/foru...log-reconstruction-filter-is-it-a-thing.2634/

The unfortunate reality is that designs with perfect transients create all sorts of issues in the frequency domain. Pick your poison.
 

andreasmaaan

Major Contributor
Joined
Jun 19, 2018
Messages
3,517
Likes
2,767
Correct on LPF, but even with analog LPF there will only be post-ringing which is said to be less noticeable as it is masked by the signal. Recent NOS R2R designs avoid the LPF altogether for this same reason and also to not change frequency/phase response, or use a very gentle LPF filter, so ringing is minimal. Technically ringing is ultrasonic so should not be audible, but dome disagree. There was a relevant discussion here:
https://audiosciencereview.com/foru...log-reconstruction-filter-is-it-a-thing.2634/

The unfortunate reality is that designs with perfect transients create all sorts of issues in the frequency domain. Pick your poison.
Thanks, this clarifies things greatly. I'm of the view that ultrasonic anything (including ringing) is - by definition - inaudible (until proven otherwise). So I'll keep my poison in the ultrasonic time domain rather than the sonic frequency domain.

My main concern here was that perhaps it was being suggested that DS DACs had some inherent disadvantage when it came to transient response in the audible range... But it seems that wasn't what was being said :)
 
Last edited:

gvl

Major Contributor
Joined
Mar 16, 2018
Messages
1,289
Likes
858
Location
SoCal
These ideas are certainly circulating around, and this is really not about DS but more about oversampling vs. non-oversampling designs, i.e. an R2R DAC can be oversampling with transient response very similar to those from DS DACs. Technically these claims have no good merit as the side effects of omitting the filters in favor of perfect transients create far worse issues that can disturb downstream equipment, and real life transients are not nearly as steep as the square impulse used for testing the transients on the test bench and because of it ringing has significantly less energy with real music and the front of real life transient doesn't suffer all that much in time domain. That said, from my experience well implemented non-oversampling DACs can sound very good, definitely better than their technical measurements may suggest, which hints at how forgiving our hearing really is.

Simply put, a properly band-limited device, such as a DAC should be, cannot have a perfect transient response by definition.
 
Last edited:

DonH56

Technical Expert
Technical Expert
Patreon Donor
Joined
Mar 15, 2016
Messages
3,166
Likes
3,866
Location
Monument, CO
The (one) problem with no output filter is that images are now large, adding ultrasonic energy that was not present in the original signal and that you likely do not want to send to your tweeters now. See e.g. https://audiosciencereview.com/forum/index.php?threads/digital-audio-jitter-fundamentals.1922/ for a brief explanation. These images also stress the output buffers and input stages of the rest of the components in the chain, adding distortion and such. I would not want a DAC with no imaging (low-pass) filter at the output. - Don
 

Similar threads

Top Bottom