• 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!

Close in jitter?

fas42

Major Contributor
Joined
Mar 21, 2016
Messages
2,818
Likes
191
Location
Australia
Okay, I've got a better handle on the picture now ... going around in circles a few times ;), talking to one or two, looking around the place ... the J-Test is useful, because aspects of what are shown appear to correlate with the perceived quality - there is no direct link, as in, close-in audio signals can be heard or not - but, an assessment of the engineering competence of the unit as it relates to the quality of playback can be made.

I aim to check out every graph easily accessible, and see if the correlation holds up ...
 

Jakob1863

Addicted to Fun and Learning
Joined
Jul 21, 2016
Messages
573
Likes
155
Location
Germany
Well, the first one is definitely wrong. Tests have been performed with music and I quoted them in the other threads:<snip>

As you see above, music selections were most definitely used (after testing with pure sine waves).

His presumption also that sinewaves show higher thresholds is just incorrect.

Not to mention the BBC paper from 1974 that uses piano music and "glockenspiel" for their evaluation.
But, although i think grimm audio could have been a bit more precise or a bit more cautious at that point, in their context (means effects of low frequency jitter on spatial impression) they might still be mainly right.
The BBC used monophonic programme material and Benjamin/Gannon explicitely emphasized the efforts done to find music that was as "tone like" as possible. Furthermore they used jitter frequencies at least two-three magnitudes abover the wander region.

The grimm pdf is dated from 2006, at that time i only new about two other examinations done in Europe on the audibility of jitter, both within university project/thesis context, but also, because following mainly the same model, using higher frequency jitter.

Ashihara in Japan used music material but random jitter.

Real program music has complex spectrum that can mask jitter better than pure tones. Here is Dolby paper on that: <snip>

And the tests above were done at much higher Jitter frequencies. At lower jitter frequencies we are talking about, masking becomes so much more powerful. Here is Dolby paper on that:<snip>

As said above (and before) its the choice of the model that matters. Obviously grimm audio´s hypothesis doesn´t fit in that model, but......

So if he has evidence that runs counter to everything we know in psychoacoustics and threshold of hearing, he needs to publish that as that will be *major news*. Just declaring it doesn't amount to anything.

Sure, as said before, doing controlled listening experiments and publish the results would have been much better. We can be sure about the publication part, but i don´t know if they had done some serious tests.

<snip> The case from audibility gets very hard and is not a domain people should enter unless they have taken such controlled tests themselves and studied psychoacoustics.

I beg to differ; you don´t have to have studied psychoacoustics to perceive a difference. To expand the models and find explanations you have to (most likely :) )
 

Jakob1863

Addicted to Fun and Learning
Joined
Jul 21, 2016
Messages
573
Likes
155
Location
Germany
<snip>
I took the initiative to run these tests. It is not hard to create these tests. Just a modicum of effort needs to go into such strong claims that run counter to full science of psychoacoustics.

I really appreciate your effort!
But i´m not sure about the "modicum" part, as i usually try to follow the traditional approach that the test equipment should be a magnitude better wrt the EUT to be safe. Sometimes i have to be glad to reach a two or three times advantage but then you´ve already be careful to calculate confounding effects.
And i´d have a hard time too (means facing similar difficulties as grimm or Mivera) to measure phase noise and wander effects in that region. My spectral analyzers aren´t sufficient for this task, so imo i´d have to do more .........
 
OP
Blumlein 88

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,524
Likes
37,057
I really appreciate your effort!
But i´m not sure about the "modicum" part, as i usually try to follow the traditional approach that the test equipment should be a magnitude better wrt the EUT to be safe. Sometimes i have to be glad to reach a two or three times advantage but then you´ve already be careful to calculate confounding effects.
And i´d have a hard time too (means facing similar difficulties as grimm or Mivera) to measure phase noise and wander effects in that region. My spectral analyzers aren´t sufficient for this task, so imo i´d have to do more .........

My cynical side has the idea this is the reason suddenly to retreat to close in jitter. It isn't easy to measure, the gear to exceed limits what tiny bit of it there is doesn't seem to be available currently. One can make good gear, use the best clocks and talk about how one hears a difference without concern measurements can call you out on it.

I do think a modicum of effort is all that is needed. The ways that jitter manifests itself are very clear cut. Either a random noise or sidebands. Once you know that then picturing what the effects of sub-picosecond jitter with periods below 10 hz is this leaves one really scratching their head as to how this could result in an audible difference. Higher, much higher, levels of jitter weren't found in listening test at frequencies of a a few hundred or thousand hz for the jitter period. Yet this low frequency stuff is far more audible? And how is it described? Much like all the other fantasy digital improvements of smoother sound, fatigue is suddenly gone, and stereo imaging snaps into place. Why and thru what mechanism could a low frequency wander so audibly effect imaging? I can't think of one.

Now all this doesn't prove anything, but the whole claim seems counter to fitting in with how all the other psychoacoustic knowledge points. So as they say extraordinary claims require extraordinary proof. The known unreliability and variability of sighted listening doesn't even begin to be adequate. So all of this is why the great efforts just to make sure properly belong to those making the claim it is so.
 

Cosmik

Major Contributor
Joined
Apr 24, 2016
Messages
3,075
Likes
2,180
Location
UK
i usually try to follow the traditional approach that the test equipment should be a magnitude better wrt the EUT to be safe.
How do they test the test equipment? :)

And if it exists and is unambiguously 'better', why not just use whatever techniques they used to make it, to build your state-of-the-art digital hi fi gear? We're not talking about diamonds and gold making the difference, just competent design and electronic modules that are probably similar to the ones in mobile phones and car sat navs.
 

Jakob1863

Addicted to Fun and Learning
Joined
Jul 21, 2016
Messages
573
Likes
155
Location
Germany
<snip>

I do think a modicum of effort is all that is needed. <snip>

To pursue the "cynical" road.....doesn´t that mean to critize others for promoting their belief (i.e. low close in noise matters much), while just maintaining another belief (i.e. a modicum of effort is all that is needed)? :)

Beside that, i agree (as stated before) that their hypothesis doesn´t fit the "usual" model and that controlled listening tests should be done to get more evidence, followed (if evidence of audibility was gained) by development of a measurement method to see whats happening at the DAC output.
 

Jakob1863

Addicted to Fun and Learning
Joined
Jul 21, 2016
Messages
573
Likes
155
Location
Germany
How do they test the test equipment? :)

And if it exists and is unambiguously 'better', why not just use whatever techniques they used to make it, to build your state-of-the-art digital hi fi gear? We're not talking about diamonds and gold making the difference, just competent design and electronic modules that are probably similar to the ones in mobile phones and car sat navs.

Thats always the most challenging part of any serious metrology effort..... ;)

But i wrote that post in the context of the "modicum effort" in listening tests wrt "wander", because i have to ensure that my listening equipment (the DAC in that case) does not mask the (might be potentially given) audible effects.
As said before, the story is usually that exchanging one box for another did make an audible difference. This audible difference will be attributed to a given technical difference, but we/one do(es) not know if that is the real cause and effect relationship. Thats why i wrote about possible side effects (maybe due to circuit imperfections). Means, it could be indeed not the "wander" that the listener detects, but for example a quirky reaction of an ic to the "wander" .
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,394
Location
Seattle Area
Not to mention the BBC paper from 1974 that uses piano music and "glockenspiel" for their evaluation.
I am not sure if you are saying that makes the test bad or good. Just in case it is the former, that is not the case at all. Glockenspiel is highly revealing track used and standardized for testing lossy audio codecs. Likewise, solo piano presents transients with quiet periods making it much easier to hear pre-echo distortion that is caused by too much quantization (compress). That quantization noise spreads before and after the transient making it audible. It is not as good as other test clips we use so it is not standardized but I am not surprised about its usage. And of course at high Wow and Flutter levels, piano is very revealing of such speed/timing changes (not a problem of course in digital).
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,394
Location
Seattle Area
But i wrote that post in the context of the "modicum effort" in listening tests wrt "wander", because i have to ensure that my listening equipment (the DAC in that case) does not mask the (might be potentially given) audible effects.
But if your DAC masks it, then it is a non-problem, right?
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,394
Location
Seattle Area
But i´m not sure about the "modicum" part, as i usually try to follow the traditional approach that the test equipment should be a magnitude better wrt the EUT to be safe.
We are able to get there with audio instrumentation. Because our bandwidth is so limited in audio, and we can transform the measurement from timing to frequency, we are able to dig deep into what is happening there. A 24-bit ADC uses in an audio measurement system with an < 2 Mhz bandwidth is a given. A source signal analyzer like the Agilent (now Keysight) that started this discussion has to operate into Gigahertz range so no option remotely exists for using 24-bit converts. Even instruments costing $100K+ would have converters that 8 or 10 bits. With their highly limited dynamic range, they would have to operate in time domain and there, their own accuracy does become an issue. Here is the keysight timing analysis noise floor that was used in the TI measurement video:

Here is the phase noise sensitivity of the Agilent E5052B in the video that Mivera/Mike post from TI:

upload_2017-5-9_8-53-7.png


Notice that at close in frequencies of 1 to 10 Hz which is the subject of these discussions, the noise level of the instrument itself as you say, rises up to - 100 db and that is using 10 Mhz bandwidth which would be insufficient to measure all oscillators. Oscillators generating higher frequencies would necessitate using 100 Mhz range where the noise floor of the instrument itself becomes -80 db. Converted to bits, that is just 13 bits of dynamic range.

Now let's compare that to the performance of last generation model of Audio Precision Audio analyzer which JA and I use (the newest is about 5 db better from what I recall). Here we are measuring an actual DAC which would show the noise floor of the instrument and the DAC. Note that this is post compensation for process gain:

upload_2017-5-9_8-57-40.png


-100 db at frequencies below 10 Hz shows us performance of around 17 bits.

To the extent we want to measure deterministic noise, we can allow process gain and there, we go way, way deep into the noise floor: (again, this is for the last generation unit that JA and I have)

upload_2017-5-9_9-0-45.png


The bottom graph shows how much process gain helps, reducing the instrument's noise floor to below -180 db or whopping 30 bits worth of noise dynamic range. This allows us to find exceptionally low level deterministic jitter components which are audibly more problematic.

So while you are right that instrumentation can limit measurements, with respect to audio band and using frequency transform, we are way, way ahead of the game as far as what is audible.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,394
Location
Seattle Area
I really appreciate your effort!
But i´m not sure about the "modicum" part, as i usually try to follow the traditional approach that the test equipment should be a magnitude better wrt the EUT to be safe. Sometimes i have to be glad to reach a two or three times advantage but then you´ve already be careful to calculate confounding effects.
And i´d have a hard time too (means facing similar difficulties as grimm or Mivera) to measure phase noise and wander effects in that region. My spectral analyzers aren´t sufficient for this task, so imo i´d have to do more .........
Let me clarify that when I said "modicum" of effort I meant for people in the industry like Bruno. For everyday person unless someone has already created the files for you, then it is impossible to do. But the claims we are examining are that of the experts put forward. They seemingly have time to go and build such circuits but can't or won't spend a day or to in Matlab to create different profile jitters to test? Or performed controlled listening tests that are documented? Both tell me that all they have done is sighted testing and with the strong bias of what they think the effects should be, they "heard" the same thing. They should know better -- certainly a lot better than putting such things in AES papers.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,394
Location
Seattle Area
One other way to intuitively evaluate the audibility of such random, low frequency jitter if you are into analog sound. Let's look at the measurement of a Linn turntable: http://www.stereophile.com/content/...power-supply-measurements#bA0gIcPIt6uEPkxt.97

Linn Turntable Jitter.png


Here an LP with a 1 Khz tone was used as the source signal for the measurement. Again, in an idealized situation we would have a single, super sharp spike at 1 Khz and nothing else.

We clearly do not have that here. There are "shoulders" or skirts around our main 1 Khz tone and that indicates random speed fluctuations or in digital lingo, "jitter." We know they are random because if they were not, they would show up as spikes as I have indicated (deterministic/periodic jitter).

Focusing back on the random, close-in jitter that is the topic of this thread, we see massive amounts of it here. The main 1 Khz tone is broadened even at levels of just -10 db! We have large amounts of it by the time we get to -50 db.

The measurements of jitter we have been showing use source signal of 10,000 Hz, not 1,000 Hz like is used above. The higher the frequency, the more pronounced the level of jitter. It is a simple matter to compensate though. To have -50 db of jitter, we need to have a timing error of 2 microseconds at 1 Khz! This is 2,000,000 picoseconds!!! Far, far higher cry than 0.5 picoseconds advocated by Mike.

As we all know, there are tons of advocates of LP playback. None complain of their stereo image being smeared even though they suffer from random jitter that is two million times higher than being advocated.

Why is that case? Reasons we have mentioned before: threshold of hearing and masking. Without these two psychoacoustic effects, no one would be able to enjoy analog formats. They have horrendous timing errors yet with good content, they are delightful to listen to. And even more so for their advocates who consider it better than digital.

The levels in analog jitter can be so high that we can just analyze them in time domain where the actual level of sound can noticeably change. More on this in another post. :)
 

fas42

Major Contributor
Joined
Mar 21, 2016
Messages
2,818
Likes
191
Location
Australia
So far using J-Test results to reveal markers is very promising - I would suggest that it might be possible to just look at that graph, alone, of a completely unknown component using a DAC, and to a far degree degree of accuracy be able to predict what the subjective impressions of the unit were, knowing nothing else. Whether that has anything to with "jitter", in the sense that people are talking about it here, I have absolutely no idea.
 

Brad

Active Member
Joined
Nov 8, 2016
Messages
114
Likes
35
I also found it curious that it's claimed that the low frequency jitter smears the stereo image.
To give a feel for the magnitude of the importance of timing, I did a geometric calculation for a time aligned speaker with tweeter and mid separated by 100mm.
If the time alignment is for 3m, with the ear at tweeter level, then moving forward 1ft (30cm) will cause a timing error of 0.55microseconds between the tweeter and mid. Surely this would smear the stereo image (ie reduced temporal coherence) more than the quoted levels of jitter. That is, people aren't that concerned about the precision of their listening distance from the speakers.
Now jitter causes amplitude fluctuations, and it could potentially smear the stereo image by uncorrelated fluctuations between the left and right speakers, that cause a time dependent left-to-right shift of the image.
So do people experience a smeared stereo image by moving their head back and forward slightly? This effect would be much greater than that possible with <ns levels of jitter
 
OP
Blumlein 88

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,524
Likes
37,057
I also found it curious that it's claimed that the low frequency jitter smears the stereo image.
To give a feel for the magnitude of the importance of timing, I did a geometric calculation for a time aligned speaker with tweeter and mid separated by 100mm.
If the time alignment is for 3m, with the ear at tweeter level, then moving forward 1ft (30cm) will cause a timing error of 0.55microseconds between the tweeter and mid. Surely this would smear the stereo image (ie reduced temporal coherence) more than the quoted levels of jitter. That is, people aren't that concerned about the precision of their listening distance from the speakers.
Now jitter causes amplitude fluctuations, and it could potentially smear the stereo image by uncorrelated fluctuations between the left and right speakers, that cause a time dependent left-to-right shift of the image.
So do people experience a smeared stereo image by moving their head back and forward slightly? This effect would be much greater than that possible with <ns levels of jitter

I don't remember where, but I have seen an estimation of how much your pulse moves your ear drums. The pulse of blood thru the hearing mechanism does move the ear drum and membrane. It was an amount of movement that was several times more than the air movement that would match up with sub-picosecond jitter levels. And as I believe I mentioned earlier, those levels of jitter would be swamped by the random brownian motion of the air. These effects just can't physically exist in a way we could hear. So you need some carefully done blind tests of significant numbers of listeners to give this much credibility. If it is true, then you have one heck of a mystery as to how it can be so in the physical world of air and sound.
 

Brad

Active Member
Joined
Nov 8, 2016
Messages
114
Likes
35
However, we do normally filter out the noise of our pulse.
People spending extended time in an anechoic chamber report that the sound of their pulse can get very loud.
 

RayDunzl

Grand Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
13,201
Likes
16,983
Location
Riverview FL

Brad

Active Member
Joined
Nov 8, 2016
Messages
114
Likes
35
I think filtered is a more apt description. The brain ignores (filters) the signal.
If it were masked, a second larger signal would be required.
The same thing happens when you take off noise cancelling headphones (you suddenly notice a lot more noise), or coloured glasses (for a short period the world looks tinted in the complementary colour to the glasses)

I should add, that the sound of the heart and breathing become loud in an anechoic chamber, only after an extended period. It does not happen immediately.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,394
Location
Seattle Area
I should add, that the sound of the heart and breathing become loud in an anechoic chamber, only after an extended period. It does not happen immediately.
Indeed. I have not felt such a sensation in brief stays in the chambers.
 
Top Bottom