isn’t noise, audible or inaudible, a distortion that would confer muddiness to SQ?
otherwise, why do Paul and PSA engineers go the extra miles to reduce it?
look at many of Paul’s videos with oodles of measurement devices
are the posters in this thread representative? Many observe issues with Mk2, many do not.
This is not the case out in the world, for those who examine specifications for their purchase decisions, for product designers, for Paul and PSA engineers, and for media reviewers.
I’ve had a busy time at work, hence not much listening in a while … BUT, the DSD Mk2 has been busily burning in the background. Just over 500 hrs in now, and turning it on tonight was just magic. I think some earlier drift in sound may have been due to playing around with the ground lift on the outputs. No such issue on turning it on tonight. Just listening to THE MUSIC!!
Over the moon with this Dac. It is waaaay beyond the Mk1 which I’ve had since it was released (as an upgrade from the Perfectwave). Congratulations @tedsmith , @Paul and the rest of the PS Audio team.
As far as I can tell the noise out of a DS Mk 1 isn’t detrimental to SQ (but, obviously, unless I heard it without I couldn’t be sure). Ted always said that each new firmware release reduced noise. maybe it did and I always ended up using a new release. Sunlight was a game changer and AFAIK Ted did something major (to do with clocking I think) and everything was so much better with Sunlight.
Now, for me, this move from the Mk 1 screen to a new screen is a major error. I won’t be able to see the volume on the Mk 2 screen so they really need to increase the size to take full advantage of the height of the display. They say the Mk 1 screen is noisy. In my system I can’t hear any difference when I turn the screen off so whatever noise it generates is being drowned out by other noise. I’d like the old touch screen back as it’s easy to use and it shows lots of information. I wish they’d have included me in the beta testing. There are lots of people saying they need a bigger display for volume (it only needs to be bigger whilst altering it as the initial touch of the remote displays volume for a few seconds). Paul has said I’ll be able to read the volume from 15 to 20 feet away - I can’t, I can read it from about 10 feet at most.
I keep the volume level at 80 on my DAC and adjust it up or down using my preamp.
I use my preamp and the Mk2 for volume adjustments. My Mk2 sounds better subtly day by day. After about 800 hours I stopped running it continuously and just leave it on without input coming in until I use it daily. (My daily use has diminished because just when I felt it had “broken in”. . . my wife walked off her job and decided to retire!)
Perhaps one reason measurements are not frequently discussed here or on other mainstream audio forums (not ASR, which is a different animal) is due to the rarity of mainstream reviews that express serious concerns about measurements?
Typically, to extent a measurement raises an eyebrow in a JA review, he’ll mitigate by describing that it really doesn’t matter much for such-and-such reason, and then say something like the product appears to be generally well engineered (or has some other redeeming trait, like it is conservatively spec’d versus its measured performance, or really anything that brings the review back to somewhat positive). It’s rare that he doesn’t end generally positively, I’d submit.
Here’s an analogy (with the caveat that analogies are never perfect and always easy to pick apart, but you’ll get the idea):
In the midwest, we don’t worry much about earthquakes. We certainly don’t discuss earthquakes in day-to-day conversation. But if the JA of earth-measuring and earthquake awareness magazine suddenly announced that she had major concerns about what midwest measurements were showing, then yeah, we’d probably be more interested and concerned. Would that mean we’d ever have a major earthquake? Hope not. Would it mean we’d have some tremors that would at least make us notice something wasn’t right? Maybe, maybe not. Who knows? But we’d be aware ot the concern at the least, and probably would be discussing…haha. So it’d be wrong to say midwesterners don’t care about earthquakes, if the proffered reason for such a view was that they almost never discuss earthquakes.
Looks like classic DSD noise hump then. Which is uncorrelated with input signal.
Curious how this would be filtered in the digital domain… lowered yes by change of modulator design (at the cost of performance inside audible band… no free lunch) but will not sure can be removed in the FPGA.
I could have it wrong though and happy to be correct
@tedsmith may share the source of this noise
I appreciate you disagree with my conclusion, you are free to do so, but measurements are infrequently discussed here and on other mainstream audio forums, and many reviews, both pro and amateur make no mention of specs.
From this one can only conclude specifications are of limited interest to audiophiles.
Or is the lack of discussion proof of overwhelming interest? 
And see the poll on this topic, “no” is winning as of this writing.
In any event, the Stereophile review’s measurements have engendered some interest in this thread.
ww, I have seen this graph before, but I cannot seem to find it. Do you have the source where you found it?
Hmmmm, maybe noise cancelling headphones for her retirement gift?
She doesn’t want to wear headphones. She wants to spend time with me doing other things than listening. I enjoy doing that as well. . . and I’ve been able to negotiate enough listening time to keep withdrawl at bay. Just some growing pains into the new “routine” but it’s going okay.
Even my little dachshund’s jealousy towards her being here all the time now is lessening. That’s a good thing.
Andreas Koch article:
https://positive-feedback.com/audio-discourse/raising-the-sample-rate-of-dsd-is-there-a-sweet-spot/
Damn, that sucks. JK, congrats on the retirement
Yes, it is currently almost 60% no 40% yes, and this is with a question phrased in such a manner to attempt to get as many yes votes as possible
Indeed
Dvorak, you asked me what specific questions I have regarding measurements and sound quality in DSD-based DACs, so I thought I’d try to give you a response. However, I don’t want to just contribute further to any overreaction to the Stereophile review.
In my search for more information about DSD (broadly) and DSD DACs (more specifically), I came across a basic summary about DSD in What Hi-Fi from last December. It’s called “What is DSD Audio?” and it doesn’t get into anything very detailed about specific DACs. It does discuss some basic differences between DSD and PCM. In the area of measurement, it states the following:
“When it comes to measurement, DSD also has issues with high levels of noise compared with PCM. Clever processing techniques allow the engineers to shift the noise above our hearing, and optimise the performance and dynamic range in the audible region. The high-frequency noise is generally filtered out.”
I read lots of discussion of different DAC designs (DSD, R2R, Ladder, Ring, NOS, etc.), but I do not see a clear consensus about which DAC design produces the most natural and realistic sound quality - and why? Are there any specific DAC measurements that correlate directly with sound quality across all DACs or do different DACs require unique subsets of measurements to accurately characterize the sound of that specific type of DAC? Or are we simply not at the point where we can accurately correlate measurements with DAC sound quality?
Thanks for any ideas you are able to share.
jk, How did you find out that the find out the ultrasonic noise is over 300 Khz? I assume that at that high of frequency it has to be due to noise shaping and requantization.
Thank you so much, reading it now, but I have to go soon. Will finish it later.
And those two are not the only ones. There are least also German magazines that do measurements on tested devices. All in all a larger audience to reckon with.
I finally got my June issue of HiFi News. For those of you who have read it, did any of you notice what I believe is a mistake in Paul Miller’s “box out” on page 47. In it he states that the 20x DSD date is noise shaped down to DSD128! It is my understanding that the MK2 converts to 20x DSD to DSD256 which is the low pass filtered to analog.
Am I confuse or has Paul Miller made a mistake in the review?