• 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!

Understanding Jitter in Digital Audio: Measurements and Listening Tests

OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,381
Location
Seattle Area
With all due respect to AES published articles, this joke of a journal has an impact factor of 1.25, this is far below any predatory journal. I would take any claim made in their articles with a grain of salt...
None of the articles I referenced were journal papers. They were all conference papers where peer review is optional.
 

Kachda

Addicted to Fun and Learning
Forum Donor
Joined
May 31, 2020
Messages
903
Likes
1,578
Location
NY
Well explained, thanks!
 

maxwell3e10

Member
Joined
Dec 29, 2020
Messages
42
Likes
23
Thanks for the overview. I didn't quite follow the purpose of 250 Hz modulation of the square wave in J-test file. Is it modulating the amplitude, frequency or phase of the 12 kHz square wave? Using a square wave as the main signal makes intuitive sense since the transitions are sharper. An interesting comparison would be an FFT in the presence of jitter for a sine wave signal, square wave signal, and square wave with 250 Hz modulation.
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,521
Likes
37,050
Thanks for the overview. I didn't quite follow the purpose of 250 Hz modulation of the square wave in J-test file. Is it modulating the amplitude, frequency or phase of the 12 kHz square wave? Using a square wave as the main signal makes intuitive sense since the transitions are sharper. An interesting comparison would be an FFT in the presence of jitter for a sine wave signal, square wave signal, and square wave with 250 Hz modulation.
One source of jitter with an SPDIF connection that really isn't an issue with other connections is when the bits in use flip. Back when the papers were written the SPDIF was pretty much the only game in town for consumers. I think it was called logic modulation. Anyway the addition of that small squarewave, its amplitude was only 1 LSB (I think), cause zero and ones to flip at the most extreme way. All of them going from zero to nearly all ones which created the most extreme possible modulation. Also if you were using a 16 bit multi-bit DAC you were also flipping all of its bits which meant switching from most current thru the DAC to least and back again in these multi-bit ladder DACs. So causing this extreme exercise of possible jitter infiltration was the purpose of the square wave at 250 hz. BTW, this is for 48 khz, it is 229.6875 hz for 44,100 hz rates. The small wave should be the sample rate divided by 192.

I'm not an expert. Don't think anything here is wrong, but if so Amir or someone can correct it.
 

dougi

Addicted to Fun and Learning
Joined
May 31, 2020
Messages
845
Likes
763
Location
ACT, Australia
From the J Dunn paper/s I believe the 250 Hz (for 48 kHz sampling) is supposed to help identify interface data dependent jitter susceptibility as opposed to random jitter. I did a comparison of my two USB interfaces on the weekend. An older USB on my Marantz NA7004 showed the 250 Hz products, hence susceptible, but my iFi nano-ISD did not, so not susceptible.
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,381
Location
Seattle Area
Thanks for the overview. I didn't quite follow the purpose of 250 Hz modulation of the square wave in J-test file. Is it modulating the amplitude, frequency or phase of the 12 kHz square wave?
It is changing the amplitude by 1 bit.
 

PeteL

Major Contributor
Joined
Jun 1, 2020
Messages
3,303
Likes
3,838
It seems like a belief originating from a malicious use of a misconception. When a cable does not have enough bandwidth for a square pulse it will spread the squares which in turn creates interference with the adjacent pulses and that can cause bit flips. Sometimes this effect is also called jitter (in communications theory it is called inter-symbol interference, jitter is mainly the timing variation).

If you heed the original Nyquist criterion (the one dealing with telegraphic transmission) your cable must have at least twice the bandwidth of your signaling rate to avoid interference and see the cable as a "flat" channel modeled by a simple gain loss, but when ISI is present and your signal to noise ratio is good there are a lot of digital techniques that will work better than a thousand dollar cable.
A digital audio signal is, in the grand world of communications, a low bandwith signal tough. Still the cables should be following standards of shielding and lenghts, but the phenomenon you describe won’t happen in audio. there might be bit flips, if noise is too high or there is too much lenght before a repeating amp impacted the amplitude, those are extreme case, but there is not a cable used in digital audio that doesn’t have a 384k bandwidth, no other digital techniques needed beside respecting the protocols standards.
I didn’t watch the video I admit, but more generally, an audio precision analyser is not the right tool to explain jitter. It can show it’s effect. An eye diagram on a scope show jitter. But in all cases, In audio, you don’t need bit flips for jitter to have an effect, all the samples decoded can be right, but if they are converted at a fluctuating rate, the resulted analog signal is distorted.
 
Last edited:

mansr

Major Contributor
Joined
Oct 5, 2018
Messages
4,685
Likes
10,700
Location
Hampshire
From the J Dunn paper/s I believe the 250 Hz (for 48 kHz sampling) is supposed to help identify interface data dependent jitter susceptibility as opposed to random jitter. I did a comparison of my two USB interfaces on the weekend. An older USB on my Marantz NA7004 showed the 250 Hz products, hence susceptible, but my iFi nano-ISD did not, so not susceptible.
That is correct. It was designed to provoke data-dependent modulation of the recovered S/PDIF clock, a common issue with older receivers. This is obviously not a problem with asynchronous USB interfaces. However, toggling (almost) all the bits at the same time can still have unwanted effects through various mechanisms, not necessarily involving clock jitter.
 

DrJayDub

Member
Joined
Jun 25, 2020
Messages
20
Likes
33
With all due respect to AES published articles, this joke of a journal has an impact factor of 1.25, this is far below any predatory journal. I would take any claim made in their articles with a grain of salt...

What is a “predatory journal”? I don’t think we have those in biology.
 

mansr

Major Contributor
Joined
Oct 5, 2018
Messages
4,685
Likes
10,700
Location
Hampshire
Gotcha! I never tried to publish without going through peer review before. Wouldn’t have even thought about it in my academic days.

What other journal for publishing peer reviewed audio science and engineering research has a higher impact factor? Also, Impact factors can be misleading sometimes....
I don't know about impact factor, but on audio topics, I'd trust papers published by IEEE more than anything from the AES.
 

EB1000

Senior Member
Joined
Jan 8, 2020
Messages
484
Likes
579
Location
Israel
The key difference between IEEE and AES is that you don't need to have any academic degree to publish in AES. All you have to do is to become a member by paying the membership fee. AES basically welcomes ignorance... The interesting fact is that even well respected audio engineers like Bruno Putzeys don't have any major IEEE publications (Bruno Putzeys is features as co author in only two IEEE papers, and about 10 AES papers).
 

audio_tony

Addicted to Fun and Learning
Joined
Feb 24, 2019
Messages
567
Likes
690
Location
Leeds, UK
@amirm In the video you indicated that when plugging in the USB connection jitter increased, I believe you said due to noise from the port.
Can an earth loop be ruled out here then? Because I've seen the same behaviour caused by a ground loop.
I don't believe that USB ports are noisy as a rule - it always seems to be ground induced noise.
 

PeteL

Major Contributor
Joined
Jun 1, 2020
Messages
3,303
Likes
3,838
With all due respect to AES published articles, this joke of a journal has an impact factor of 1.25, this is far below any predatory journal. I would take any claim made in their articles with a grain of salt...
The thing is, beside the impact factor, what do you find erroneous with his research and experiments? Sure it’s open to discussion, that’s one men finding. But, hey, it’s audio reproduction folks, not the formula of the Covid vaccine. we have to put things in perspective, not everything is a major scientific breaktrough, that deserve an article in science publications and peer review. You can take it for what it is and if some bits of it is inexact or not fully complete you can of course flag it. But blanket statements to the effect that it can’t be trusted is not really helping. Some people do research, and it helps the state of the art. If we need a comitee to confirm every times someone want to spread knowledge and studies, we wouldn’t have the advancement we have with audio fidelity products. It still better to use this as a reference than a youtube video...
 

scott wurcer

Major Contributor
Audio Luminary
Technical Expert
Joined
Apr 24, 2019
Messages
1,501
Likes
2,821
The key difference between IEEE and AES is that you don't need to have any academic degree to publish in AES. All you have to do is to become a member by paying the membership fee. AES basically welcomes ignorance... The interesting fact is that even well respected audio engineers like Bruno Putzeys don't have any major IEEE publications (Bruno Putzeys is features as co author in only two IEEE papers, and about 10 AES papers).

I've done both neither charged any publication fees. My AES conference presentation was rejected as a journal article, the circuit presented eventually became a multi-million dollar product that is still in production 30yrs. later. It also got me an invitation from the LIGO team to visit their facility in appreciation for the socket it filled (as well as Bruce Hofer of AP). BTW there have been plenty of dropout presenters in the IEEE.
 
Last edited:
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,381
Location
Seattle Area
The key difference between IEEE and AES is that you don't need to have any academic degree to publish in AES.
Where are you reading the requirement for a degree for IEEE? As far as I know, they accept equivalent work experience in lieu of a degree.

As to AES membership, you have to have three sponsors to get accepted as a member. It is not as simple as just paying the money.

I personally don't like IEEE because they take free papers from authors and then charge an arm and a leg for access to them. They think everyone is a major corporation and can pay their fees.
 

mansr

Major Contributor
Joined
Oct 5, 2018
Messages
4,685
Likes
10,700
Location
Hampshire
I personally don't like IEEE because they take free papers from authors and then charge an arm and a leg for access to them. They think everyone is a major corporation and can pay their fees.
As do most of the academic publishers, sadly. Personally, I have access to IEEE but not AES.
 
Top Bottom