Low-Jitter USB: Dan Lavry, Michael Goodman, Adaptive, Asynchronous
Nov 23, 2010 at 10:53 AM Post #136 of 166

Bmac

100+ Head-Fier
Joined
Feb 23, 2009
Posts
300
Likes
14

Quote:
Originally Posted by Bianci1969
 
Synchronous USB only works at 48khz. Not a practical solution for all that pesky 44.1khz music out there.
 
 Bianci
 


Really? So all the synchronous DAC's in the world don't work with CD audio? That's news to me (and probably several DAC manufacturers and their thousands of clients).
 
Nov 23, 2010 at 11:36 AM Post #137 of 166

Bmac

100+ Head-Fier
Joined
Feb 23, 2009
Posts
300
Likes
14


Quote:
Three things:
 
1) The measurement system that Stereophile uses was designed specifically to measure the jitter added by the S/PDIF interface. When used in other applications it can only show gross defects. It has a measurement limit of 120 psec for 16 bit data (which is how it is normally used). Furthermore there have been several changes to the hardware over the years. Quoting numbers from Stereophile measurements as a way to "prove" the audible superiority of a product is just plain wrong.
 
2) If "good enough" is good enough for you, that's great. You can save yourself a lot of money by purchasing low-cost products that are "good enough". We are interested in the highest quality products and making things that are better than anyone has ever made before. They're not for everybody, and they may not be for you.
 
3) You haven't even tried (perhaps wisely) to dispute my point that (if all else is equal) that an asynchronous DAC will have lower jitter than an adaptive DAC, simply because a fixed oscillator clock will have lower jitter than a variable oscillator (eg, PLL). So if you are happy with higher jitter, more power to you. I am not. Furthermore, Mr. Goodman's post contained falsehoods that demanded correction.

In response to point 1: I wasn't quoting the numbers to prove audible superiority, and I don't think that was implied either. For $2,500 I would fully expect your DAC to sound subjectively better than a $400 competitor. I was however quoting Stereophile's numbers because they are the only published jitter measurements that I know of that are third party, reasonably reliable and which contain at least a few of both types of USB DAC's. Yes the measurement hardware will change over time, but I believe all the DAC's I mentioned were measured within a short period of time using the same gear. So I submit that the Stereophile measurements still have some validity with regard to jitter until something more reliable is available.
 
In response to point 2: I never said that "good enough" was good enough. Building products that are technically better is an admirable pursuit. Having said that, I haven't actually seen any proof yet that well-engineered asynchronous DAC's in general have lower jitter at the analog outputs than well engineered synchronous DAC's. But assuming they do have lower jitter, the other problem is that again there is no proof that a product with 50ps of jitter can be discerned by human ears from a product with 120ps of jitter. I don't think they can. The asynch camp has been very vocal about the fact that their products sound better because they have lower jitter. That seems to be the only claim for audible superiority. I think the onus lies on the proponents who are very vocal about their superiority to prove that their products are in fact technically superior, and then that the technical superiority is actually audible by human ears. Neither of those criteria has been satisfied yet.
 
In response to 3: I'm not an engineer, so I won't attempt to dispute the point that all other things being equal that asynchronous will have lower jitter than synchronous. It very well may be fact. Having said that though, there are an infinite number of ways and parts to design a DAC with so in the real world all other things are probably never equal. Every DAC takes it's own approach; what matters is the end result. From what I've seen anyway, synchronous DAC's can be just as capable of low- jitter output as asynchronous.
 
Nov 23, 2010 at 12:27 PM Post #138 of 166

nick_charles

Headphoneus Supremus
Joined
Feb 26, 2008
Posts
3,180
Likes
335


Quote:
Bmac said:
In response to point 1: I wasn't quoting the numbers to prove audible superiority, and I don't think that was implied either. For $2,500 I would fully expect your DAC to sound subjectively better than a $400 competitor.
 
Since it is quite common for manufacturers of exotic equipment to just drop in something modest in a shiny box and add $3500 to the price I would hesitate to expect as a general rule a strong correlation between price and quality, the laws of physics are indifferent to price tags and circuit design more important to quality.
 
the other problem is that again there is no proof that a product with 50ps of jitter can be discerned by human ears from a product with 120ps of jitter. I don't think they can.
 
There is no reliable evidence to suggest that when the only variable altered is jitter that any difference in jitter magnitude below absurd levels will be noticed. There is much chat about how difficult it is to measure jitter correctly, how to characterize it and where to measure it and so on and some commentators such as Bob Adams (Analog Devices) suggests why bother as jitter will appear in downstream distortions such as IMD. High Jitter is generally correlated with high IMD. We can't agree on what type/format is more common or more damaging or on how and where to measure it and on the utility of tests of jitter audibility but we still know somewhow magically that it is a big problem, but we cannot back this up with any reliable listening tests showing that it is a problem, makes me wonder why we worry about it at all
confused.gif

 
 

 

 
Nov 24, 2010 at 5:56 PM Post #139 of 166

Bianci1969

New Head-Fier
Joined
Oct 22, 2010
Posts
15
Likes
16
Hello Bmac-
 
To quote the Well Tempered Computer..."

Synchronous

The clock is directly derived from the 1 kHz frame rate. There is a PLL that takes in the start of frame signal and generates a clock. Using this scheme its rather difficult to generate 44.1, but very easy to generate 48 kHz. This is a primary reason why many early USB audio devices only supports 48 kHz, they used this mode"
 
I guess it's not impossible, but impractical. If you could, tell me what USB dac's are running synchronous usb at 44.1khz. I'm not aware of any.
 
Bianci
 
Nov 26, 2010 at 7:55 PM Post #141 of 166

Bianci1969

New Head-Fier
Joined
Oct 22, 2010
Posts
15
Likes
16
 
if you divide 12MHz by 44.1kHz you get 11.2896...MHz. So why bother using synchronous when the clock rate vs the standard sampling rate of 44.1kHz are not mathematically compatible? 48Khz is nice and tidy for synchronous but why bother implementing this format if 99.999999999999999999999999999999999% of all music sold is 44.1?
 
I didn't see any mention of what type of USB interface Nuforce uses on this piece. But if the measurements from a recent CD player review are any indication of their engineering abilities I'll pass...
 
"This is one of the highest levels of jitter I have measured, and about 20 times higher than in the best products I have measured." - Atkinson
 
http://stereophile.com/content/nuforce-cdp-8-cd-player-measurements
 
Nov 26, 2010 at 9:24 PM Post #142 of 166

nick_charles

Headphoneus Supremus
Joined
Feb 26, 2008
Posts
3,180
Likes
335


Quote:
 
But if the measurements from a recent CD player review are any indication of their engineering abilities I'll pass...
 
"This is one of the highest levels of jitter I have measured, and about 20 times higher than in the best products I have measured." - Atkinson
 
http://stereophile.com/content/nuforce-cdp-8-cd-player-measurements


that is a lot of jitter...and yet the listening reviewer thought it was a good CD player, either it is another one of Stereophile's many deaf reviewers or ...perhaps jitter just does not make that much diffference ?
 
 
Nov 26, 2010 at 10:04 PM Post #143 of 166

HeadphoneAddict

Headphoneus Supremus
Joined
Aug 18, 2007
Posts
17,450
Likes
832


that is a lot of jitter...and yet the listening reviewer thought it was a good CD player, either it is another one of Stereophile's many deaf reviewers or ...perhaps jitter just does not make that much diffference ?
 





I have a CDP-8 which sounds better than my Apoge mini-DAC. Nuforce found the cause of the jitter and fixed it for future shipments, and offers the fix to current owners. I'll send mine in for the update; but apparently low frequency jitter is harder to hear and doesn't wreck the sound. Everyone who heard mine at RMAF was impressed with it.
 
Nov 27, 2010 at 7:10 PM Post #144 of 166

mgoodman

Member of the Trade: CEntrance
Joined
Sep 14, 2009
Posts
817
Likes
256
@ CHanson,
 
Chris - You need to stop accusing me of falsehood and instead re-read your own argument: "(if all else is equal) that an asynchronous DAC will have lower jitter than an adaptive DAC".
 
1) What specifically is the "Else" that's being equal? Consider that the "Else" in our Adaptive circuits keeps winning industry recognition, magazine awards and customer satisfaction. "All else being equal" is a theoretical argument that's disconnected from reality.
 
2) It's not enough to say that "in theory" your approach is better. Everyone on this forum knows that we don't live in a world of theory. We live in a real world, where practical design is what makes or breaks the product. Put a bad clock in your Asynchronous implementation and you've just made the worst DAC ever, despite using your "theoretically correct" approach.
 
So let's stop hiding behind purely theoretical arguments - they don't get you anywhere. Let's get practical and focus instead on making the best products we can make.
 
Nov 28, 2010 at 12:45 AM Post #145 of 166

regal

Headphoneus Supremus
Joined
May 27, 2006
Posts
3,645
Likes
18
An abacus can do addition and substraction,  in theory a calculator can do the same with better speed and accuracy.  Yet the abacus makers didn't have the programmers and engineers with the know how to make calculators,  so they went out of business.   Why don't you hire a few software engineers and develop an asynchronous protocol to at least compare to your mighty adaptive devices,  companies that don't explore new developments typically don't last long.
 
Nov 28, 2010 at 8:08 PM Post #146 of 166

Bianci1969

New Head-Fier
Joined
Oct 22, 2010
Posts
15
Likes
16
Mr. Goodman,
 
I guess the question is this-. what implementation is better, Asynchronous or Adaptive? If you could figure out the code to Asynchronous would you be using it? If Gordon Rankin or DCS were offering you a license would you take it? Or, do you stand behind Adaptive as a superior method for transferring audio over USB?
 
Bianci
 
Nov 29, 2010 at 1:58 PM Post #147 of 166

Bmac

100+ Head-Fier
Joined
Feb 23, 2009
Posts
300
Likes
14

Quote:
Why don't you hire a few software engineers and develop an asynchronous protocol to at least compare to your mighty adaptive devices,  companies that don't explore new developments typically don't last long.

 
Is this a serious question or just bad trolling?

Quote:
Mr. Goodman,
 
I guess the question is this-. what implementation is better, Asynchronous or Adaptive? If you could figure out the code to Asynchronous would you be using it? If Gordon Rankin or DCS were offering you a license would you take it? Or, do you stand behind Adaptive as a superior method for transferring audio over USB?
 
Bianci

Bianci, you are aware that the synchronous USB protocol that Mr. Goodman developed for Centrance is licensed to other manufacturers and is used in some well respected (and measuring) DAC's, right?
 
Nov 29, 2010 at 9:22 PM Post #148 of 166

audioengr

Member of the Trade: Empirical Audio
Joined
Aug 15, 2004
Posts
1,092
Likes
17
 
Quote:
 that is a lot of jitter...and yet the listening reviewer thought it was a good CD player, either it is another one of Stereophile's many deaf reviewers or ...perhaps jitter just does not make that much diffference ?
 

 
This is like so many reviews.  You dont know the quality of the reviewers system.  Listing the components tell you nothing really.  The only way to know is to be there and play a sample track.  They typically dont do shootouts so you dont know what they are comparing it to.  Some online reviewers that dont care about stepping on toes actually do shootouts.  I like these guys because I'm not afraid of losing.
 
There are dozens of preamps out there that create so much noise and distortion that any benefits of lower jitter are lost in the noise.  Jitter can also be euphonic, like "tubiness", having a masking effect on other system nasties, actually a positive thing.  This is definitely going down the garden path however.....
 
The reviewers are not deaf, but their systems are sometimes crippled IME.
 
Steve N.
Empirical Audio
 
Nov 29, 2010 at 9:35 PM Post #149 of 166

audioengr

Member of the Trade: Empirical Audio
Joined
Aug 15, 2004
Posts
1,092
Likes
17
 
Quote:
  if you divide 12MHz by 44.1kHz you get 11.2896...MHz. So why bother using synchronous when the clock rate vs the standard sampling rate of 44.1kHz are not mathematically compatible? 48Khz is nice and tidy for synchronous but why bother implementing this format if 99.999999999999999999999999999999999% of all music sold is 44.1?

 
You need to check your math.  If you divide 11.2896MHz by 256 you get 44.1kHz.  In Async systems multiples of 44.1 and 48 are used for the two clocks.  Even multiples.
 
In one CEntrance Adaptive interface a 6MHz clock is multiplied up to 48MHz.  This is then used to synthesize all of the sample-rate frequencies.  You can easily look this up in the datasheets.
 
Steve N.
Empirical Audio
 
Nov 29, 2010 at 10:15 PM Post #150 of 166

bigshot

Headphoneus Supremus
Joined
Nov 16, 2004
Posts
24,875
Likes
5,551
Location
A Secret Lab

Users who are viewing this thread

Top