Jitter test

Vincent Kars

WBF Technical Expert: Computer Audio
Jul 1, 2010
860
1
0
There is a strong tendency in the audiophile community to minimize processing on a dedicated audio PC as much as possible.
There is a rationale, any process not running cannot kick in and causes a dropout.
However, the claims are much bolder. Killing system services, disabling network drivers, swapping a HD by a SSD, etc. etc all have a profound impact on sound quality.
The idea is that all electrical activity has a negative impact on sound quality. Hence, if you minimize this sound quality should improve.

Archimago did some interesting experiments.
He measured the jitter using adaptive & asynchronous USB and Toslink.
Both under (unspecified) normal conditions and with the CPU and GPU running at 100%.
His jitter test most of all demonstrate the very low impact of a 100% CPU/GPU load on jitter.

http://archimago.blogspot.ca/2013_03_01_archive.html
 

Orb

New Member
Sep 8, 2010
3,010
2
0
Hmm,
it is interesting that Paul Miller's instrumentation has measured differences with USB cables (ironically one of the best he found was a cheap cable but I think no longer available) and implementation of USB DACs vs "bridge" USB to SPDIF products.
Personally I would never be without the well designed USB to SPDIF type product as they isolate the various jitter so much better it seems; Halide-Musical Fidelity-etc.
Two considerations I thought, electrical interference and separately resource-processor overheads to filter algorithms.

Cheers
Orb
 

dallasjustice

Member Sponsor
Apr 12, 2011
2,067
8
0
Dallas, Texas
What if the problem to be addressed isn't jitter? A few respected engineers like Demian Martin believe that electrical noise from the computer "rides" on the signal line. The digital signal is just an analog representation of 1s and 0s. Even if the timing isn't affected, computer noise comes through and can negatively affect some DACs. I'm not saying it's the same with every DAC, but jitter isn't really the reason people go to great lengths to reduce electrical noise on the server side.
 

dallasjustice

Member Sponsor
Apr 12, 2011
2,067
8
0
Dallas, Texas
Archimago measures the analog out of the DAC.
If some noise creeps into the DAC and affects the conversion, I expect it to be measurable at the analog out.
I am not convinced this guy knows much about jitter. Look at this page:
http://archimago.blogspot.ca/2013/03/measurements-adaptive-aune-x1.html
He's got a picture of an rca terminated "digital cable" and refers to it as a "proper coaxial" cable. He's trying to measure difference in cables but doesn't understand that rca can NEVER be 75 ohms; only some BNC can be 75 ohms, not to mention the cable itself is unlikely 75 ohms.

The tests on digital interface reflections were done many years ago by Pat from Analog Research and he's posted all of his findings online. It's easy to see that digital interfaces DO induce jitter, if you know what you are looking for. Many others know about this phenomenon like the folks at Berkeley with their high return loss digital output on their Berkeley Alpha USB. I think even Steve Nugent has written on the subject and posted some test results.

I am fine with skepticism, but I don't find these tests very well informed.
 

jkeny

Industry Expert, Member Sponsor
Feb 9, 2012
3,374
42
383
Ireland
Have to agree with DJ here. I believe Amir has posted that using the 16/44 Dunn J-Test signal is not appropriate for measuring jitter on a USB DAC because it was designed specifically to stress SPDIF transports which use PLLs & are effected by intersymbol interference (ISI). This is not suitable for USB protocol.

I'm not at all convinced by his measurements or listening & statements like this make me even more convinced that he is working on a "belief system"
I do not believe it is audible with music unless extreme like maybe >10ns. It's also worth remembering that research into the threshold of jitter audibility decreases with higher frequencies; but this also correlates with human hearing losing sensitivity the higher up we go - especially as we get older (therefore, sensitivity to jitter should decrease with age). These days, well engineered "high fidelity" gear should not be even close to showing 10ns jitter. Likewise, software player programs should be capable of "bit-perfect" output without difficulty or cause timing errors.
 

dallasjustice

Member Sponsor
Apr 12, 2011
2,067
8
0
Dallas, Texas
I think its an interesting topic though. Ive never seen any credible evidence that a player can reduce jitter over asynchronous USB. Many have claimed it; eg. JPLAY, CMP cplay. Cplay claims they've proven lower jitter but I can't find the proof of a comparison to other software on their website. I can say that jplay does sound better than jriver; But not a tremendous difference. I think the major benefits to software is the potential to use a variety of filters, upsampling, dither; HQplayer.
 

jkeny

Industry Expert, Member Sponsor
Feb 9, 2012
3,374
42
383
Ireland
I think its an interesting topic though. Ive never seen any credible evidence that a player can reduce jitter over asynchronous USB. Many have claimed it; eg. JPLAY, CMP cplay. Cplay claims they've proven lower jitter but I can't find the proof of a comparison to other software on their website. I can say that jplay does sound better than jriver; But not a tremendous difference. I think the major benefits to software is the potential to use a variety of filters, upsampling, dither; HQplayer.
I too can hear differences between Jplay & other players. I can also hear differences between some USB cables - it just proves that there is a lot more we need to learn about what/how to measure.
You mentioned Pat Giacommo (aka Jocko Homo) from ART & his expertise & knowledge of the RF measurements in audio. I recently read this from him ((quoted from a couple of different forums) "If you are into "imaging" and "three dimensionality", you must have low 1/f jitter....... Allan Deviation is the best way to characterize clocks."
Measuring this sort of jitter is pretty difficult as is measuring "noise modulation" that Opus11 talks about & says he can hear in recordings & playback using certain types of DACs.
 

dallasjustice

Member Sponsor
Apr 12, 2011
2,067
8
0
Dallas, Texas
Pat is a friend. Ive visited with him many times standing in front of his house. He wont let me come in. :) I wouldn't dispute anything he says. If anyone knew how to measure jitter and had the right gear to do it, it would be Pat. Btw, he claims he's not the infamous Jocko Homo, but I don't believe him.
 

jkeny

Industry Expert, Member Sponsor
Feb 9, 2012
3,374
42
383
Ireland
Pat is a friend. Ive visited with him many times standing in front of his house. He wont let me come in. :) I wouldn't dispute anything he says. If anyone knew how to measure jitter and had the right gear to do it, it would be Pat. Btw, he claims he's not the infamous Jocko Homo, but I don't believe him.
Well, I'm not so sure about not disputing anything he says ( the Jocko Homo denial, for instance) but in matters RF, he has lots of knowledge & experience.

In his other alterego, Gen Cartman Lee, here is what he also says on the same subject
OK, bub.................you say you have measured noise floor. Sure, but how low? I know you have not gone low enough.

"How low is low enough?"

Below 0.1 Hz.

Yes, that is right, well below 1 Hz. You didn't, did you?

Of course not, because it is not easy. You would need to do it over a several hour time period. And then, it would need to be done several times, to get a reliable answer.

The Devil has given you the "1/f noise" clue enough times that it should be obvious. And since there is no Maxwell Demon, it can not tell signal (music) from noise. It has to work where there is no signal!

How many more clues do you need?
 

dallasjustice

Member Sponsor
Apr 12, 2011
2,067
8
0
Dallas, Texas
Yes. Very few folks have the gear to measure down that low. I think MSB measures down to 1hz. Here's a recent measurement he took of a sample Russian crystal.
c1.jpg
 

jkeny

Industry Expert, Member Sponsor
Feb 9, 2012
3,374
42
383
Ireland
Yea, looks pretty good (got a manufacturer/model for it?)but phase noise plots are frequency dependent i.e this is taken at 11.xxMHz - it could be very different at 22.XXXMHz
According to him, the more reliable figure to characterise clocks seems to be Allan Deviation.
 

Julf

New Member
Nov 27, 2011
613
0
0
Amsterdam, The Netherlands
The tests on digital interface reflections were done many years ago by Pat from Analog Research and he's posted all of his findings online. It's easy to see that digital interfaces DO induce jitter, if you know what you are looking for.

Sure, there will always be some jitter. But many of the latests DACs employ buffering and reclocking, so that input jitter doesn't affect the output. You will still have jitter from the DAC output clock. The question is: will it be audible? When we see an amp with 0.01% THD, we don't go "hey, look, there is DISTORTION!". There seems to be surprisingly little solid data available on what amount of jitter (and what kind of jitter) is actually audible.
 

Julf

New Member
Nov 27, 2011
613
0
0
Amsterdam, The Netherlands

jkeny

Industry Expert, Member Sponsor
Feb 9, 2012
3,374
42
383
Ireland
Under controlled, double-blind conditions?

No, I never had the money or time to hire the necessary expertise to do a statistically, scientifically rigorous DBT, have you?

Or if you mean did I use the toy version of going out of the room & having someone change (or not) the software player & me identifying if there was a change when back in the room (blind) - then yes.

But this would hardly qualify as controlled, DBT & wouldn't be an acceptable test by anyone who is serious about such tests. So I guess this isn't acceptable to you, or is it?
 

jkeny

Industry Expert, Member Sponsor
Feb 9, 2012
3,374
42
383
Ireland
Sure, there will always be some jitter. But many of the latests DACs employ buffering and reclocking, so that input jitter doesn't affect the output. You will still have jitter from the DAC output clock. The question is: will it be audible? When we see an amp with 0.01% THD, we don't go "hey, look, there is DISTORTION!". There seems to be surprisingly little solid data available on what amount of jitter (and what kind of jitter) is actually audible.
Sure, I know the theory but in practise there still seems to be audible differences between high & low jitter sources when using such players. So, perhaps, as DJ said it ain't just jitter that is at the heart of the matter - maybe the 1/f noise is varying - maybe it's noise modulation, ala Opus111, maybe it's something else?
 

Julf

New Member
Nov 27, 2011
613
0
0
Amsterdam, The Netherlands
But this would hardly qualify as controlled, DBT & wouldn't be an acceptable test by anyone who is serious about such tests. So I guess this isn't acceptable to you, or is it?

What I have done is I have had 2 computers, with identical setups, being served by the same network storage, playing in parallel into a computer-controlled switch, and I have had a program first playing from one, then from the other, and then from a randomly chosen one, where I had to then select "A" or "B" to tell if the third one was the same as the first or the second one. Not scientifically rigorous enough to prove a scientific theory with, but quite a bit better than sighted or single-blind tests - and not very expensive or difficult to do. Tedious, yes...

Originally I started out with one computer with two (similar) audio interfaces, and both programs running on the same computer, but there were too many programs that didn't work nicely together. Another challenge was starting the programs at the same time - fortunately I have access to a web application test suite with the capability of "driving" an application based on an automatic (and remote-controlled) script - and I made sure the switch-overs included two seconds of silence, to minimize the detectability of tiny timing variations. Not perfect, but good enough for my purposes.
 

About us

  • What’s Best Forum is THE forum for high end audio, product reviews, advice and sharing experiences on the best of everything else. This is THE place where audiophiles and audio companies discuss vintage, contemporary and new audio products, music servers, music streamers, computer audio, digital-to-analog converters, turntables, phono stages, cartridges, reel-to-reel tape machines, speakers, headphones and tube and solid-state amplification. Founded in 2010 What’s Best Forum invites intelligent and courteous people of all interests and backgrounds to describe and discuss the best of everything. From beginners to life-long hobbyists to industry professionals, we enjoy learning about new things and meeting new people, and participating in spirited debates.

Quick Navigation

User Menu

Steve Williams
Site Founder | Site Owner | Administrator
Ron Resnick
Site Co-Owner | Administrator
Julian (The Fixer)
Website Build | Marketing Managersing