Jitter vs no-Jitter

[bibo01]

Jitter vs no-Jitter

 

This picture is a Jitter test of DAC no.1. Jitter is so low at this window that practically it cannot be measured.

 

 

 

This picture is a Jitter test of DAC no.2. Jitter is high and over 200ps.

 

 

Question -

Listening to two equal tracks - "A1" and "A2" - recorded passing through each DAC, are we able to recognize DAC no.1 with less jitter?

 

In the pool options are:

1) track "A1"

2) track "A2"

3) I am not sure to hear differences.

 

This is a link to download the two tracks in question: dropcanvas - instant drag and drop sharing - canvas view

The answer is in a ZIP password protected file. The answer will be given in 10 days.

 

Thanks for participating

[徐中銳]

Incidentally, have you tried DSD vs PCM, Which sounds the best to your ears of these free same source files?

[sandyk]

test ascolto.zip is unsafe to download and was blocked by Smart Screen Filter.

(Windows.)

 

Irrespective, in my experience, as well as that of Cookie Marenco from Blue Coast Records ( a major DSD proponent) using COMPRESSED Zips degrades sound quality.

Neither do you say what format the files are in. (.aiff , .wav etc.)

[bibo01]

(Windows.)

 

Irrespective, in my experience, as well as that of Cookie Marenco from Blue Coast Records ( a major DSD proponent) using COMPRESSED Zips degrades sound quality.

Neither do you say what format the files are in. (.aiff , .wav etc.)

The audio files are in FLAC format.

The ZIP file only contains the "correct" answer. If your anti-virus gave a warning on that file, it's just a "fail safe" (or however it is called).

[christopher3393]

The audio files are in FLAC format.

The ZIP file only contains the "correct" answer. If your anti-virus gave a warning on that file, it's just a "fail safe" (or however it is called).

 

famous...last...words...

[elcorso]

(Windows.)

 

Irrespective, in my experience, as well as that of Cookie Marenco from Blue Coast Records ( a major DSD proponent) using COMPRESSED Zips degrades sound quality.

Neither do you say what format the files are in. (.aiff , .wav etc.)

 

Hi Alex,

 

It's uncompressed ZIP and in FLAC format. Downloaded without problems. Maybe the only virus is jitter in one of them . Didn't listened yet.

 

Regards,

 

Roch

[bibo01]

BTW, I would like to let you know that I posted the same pool also on WTB and NextHardware forums to possibly get a large number of testers.

 

If I have forgotten some details, forgive me as this is my first pool, but please do not hesitate to ask.

[徐中銳]

Okay, let's recall these PS Audio uploads by Paul McGowan :

Jitter and why it matters, Consulting PS Engineer Ted Smith explains to us why jitter matters. How is it jitter makes so much difference in audio while not quite as important in imaging and other sciences.

We all know jitter, in digital audio, isn't something we want. But why? How does it sound? How does it affect the way our systems perform and what does it do to the music? PS Audio consulting digital expert, Ted Smith, gives us a clear picture of what damage jitter does with the music.

 

Ted Smith explains clocks and their importance in the audio chain. All audio systems run on clocks and the quality of those clocks contributes to one of the major problems in sound, jitter. How do expensive, specialized clocks like rubidium devices affect the jitter? You might be surprised at the answer.

[sandyk]

The audio files are in FLAC format.

The ZIP file only contains the "correct" answer. If your anti-virus gave a warning on that file, it's just a "fail safe" (or however it is called).

 

Windows blocked the Download, not my Norton Anti-Virus.

[bibo01]

Windows blocked the Download, not my Norton Anti-Virus.

 

On Windows 7 I downloaded it fine. Perhaps you have to check your settings. Anyhow, the file is safe; perhaps it is something to do with the site.

[sandyk]

On Windows 7 I downloaded it fine. Perhaps you have to check your settings. Anyhow, the file is safe; perhaps it is something to do with the site.

 

Windows 8.1 blocked it, which is unusual, as it's normally Norton that does such things, and there appears to be no easy way to overcome it without too much hassle. The blurb about why it was blocked did not give a way to bypass it. Without the Zip file, which appears to be superfluous at this point in time , there would have been no problems.

Whether or not, the file with less Jitter is preferred, will come down to the actual system itself, as surprisingly, many people appear to prefer the effects of a little added Jitter.

[bibo01]

Perhaps you can download it from a different system.

[esldude]

I keep getting told I am between #10 and #15 in a download queu, then it starts to download after I reach #1. Then each time it downloads about one third of the file and fails. Maybe if someone will host it elsewhere or set up a bittorrent for you. The torrent would be very nice.

[esldude]

Finally got the file to download. Alex, if you will PM me your email address, I can filemail it to you. I am not at my home machine so don't have your address handy.

 

Also after two quick listens I voted in the poll, but voted backwards. I thought we were trying to identify the jitter, I guess we were trying to identify the low jitter dac. Oops. My vote should have been reversed.

[JohnSwenson]

What jitter test?

 

I see two spectrums, the second has a higher "noise floor" and some spikes. These spikes are not arranged in any sequence that would indicate they were caused by jitter, thus I do not see how they can actually be showing any difference in jitter.

 

Yes there is more "garbage" on the output of DAC2, but that does not automatically mean it was caused by jitter.

 

How was the 200ps of jitter determined? Was it measured by actually using a jitter analyzer on the clock at the DAC chip or by some interpretation of the above graph? If the latter it is almost certainly not 200ps of jitter but something else causing what is seen in the graph.

 

Given that the main signal is a little less than 25KHz this is NOT a traditional JTEST, so what is the input?

 

The output of the two DACs are sllightly different, but it does not look like it is caused by 200ps of jitter.

 

It sounds like the outputs of the two DACs were fed through an ADC to get the files that are linked here, for the test to have much validity that ADC should have lower noise and jitter than either DAC, do you know the specs on the ADC used?

 

Sorry for the rant, I'm just seeing a lot of "tests" these days that are purported to measure jitter, but are really measuring something else. I think jitter IS important, but if your doing jitter tests I think you actually need to measure jitter not something else.

 

John S.

[sandyk]

Finally got the file to download. Alex, if you will PM me your email address, I can filemail it to you. I am not at my home machine so don't have your address handy.

 

Also after two quick listens I voted in the poll, but voted backwards. I thought we were trying to identify the jitter, I guess we were trying to identify the low jitter dac. Oops. My vote should have been reversed.

 

Thanks Dennis, but Windows seems likely to reject it again. I found a way to disable Windows Smart Screen Filter, but the Editor of that forum added that he didn't recommend doing so.

I will pass this time around

Alex

[esldude]

What jitter test?

 

I see two spectrums, the second has a higher "noise floor" and some spikes. These spikes are not arranged in any sequence that would indicate they were caused by jitter, thus I do not see how they can actually be showing any difference in jitter.

 

Yes there is more "garbage" on the output of DAC2, but that does not automatically mean it was caused by jitter.

 

How was the 200ps of jitter determined? Was it measured by actually using a jitter analyzer on the clock at the DAC chip or by some interpretation of the above graph? If the latter it is almost certainly not 200ps of jitter but something else causing what is seen in the graph.

 

Given that the main signal is a little less than 25KHz this is NOT a traditional JTEST, so what is the input?

 

The output of the two DACs are sllightly different, but it does not look like it is caused by 200ps of jitter.

 

It sounds like the outputs of the two DACs were fed through an ADC to get the files that are linked here, for the test to have much validity that ADC should have lower noise and jitter than either DAC, do you know the specs on the ADC used?

 

Sorry for the rant, I'm just seeing a lot of "tests" these days that are purported to measure jitter, but are really measuring something else. I think jitter IS important, but if your doing jitter tests I think you actually need to measure jitter not something else.

 

John S.

 

24 khz test signal on a 96 khz sample rate is one you could use for jitter.

 

I do agree the spikes I see don't look symmetrical or like jitter.

[esldude]

Thanks Dennis, but Windows seems likely to reject it again. I found a way to disable Windows Smart Screen Filter, but the Editor of that forum added that he didn't recommend doing so.

I will pass this time around

Alex

 

Okay Alex. I btw, being away from home was using a Win 8.1 laptop. I haven't turned off any filtering and it didn't complain at me. I could send you the two FLAC's if you wish already unzipped.

[sandyk]

Dennis

John Swenson's post wasn't there before I finished posting my earlier reply. I also had concerns about the validity of this test, which is why I didn't try too hard to find a way to bypass Windows Smart Screen Filer, which is aimed mainly at Phishing Sites etc. and became lukewarm about the whole thing, which should be obvious from my later replies.

Were both of the DACs identical in both the digital and analogue areas ? (even then...) If not the whole test is flawed as far as I am concerned.

 

Alex

[bibo01]

To John Swenson and others,

 

Almost all noise signals present in graphic no.2, analyzing them one by one, are simmetrical with a jitter frequency.

Testing setup is the same for both converters, but one of them had a damaged clock.

 

Maximum spike elevation reaches -108dB.

Minus -3dB of the fundamental we have a value of -105dB.

Now we can calculate jitter as: 10^(-105)/20*2/pi/24000*10^12 = 149,16ps

To 149ps we must add other spikes (periodic and stochastic level), so we can claim that all together the jitter value is above 200ps.

 

EDIT: for example (one of many) bands with highest peaks are at frequency 13.5kHz and 34,5kHz. It means a frequency of symmetrical jitter of 10,5kHz.

[bibo01]

Okay Alex. I btw, being away from home was using a Win 8.1 laptop. I haven't turned off any filtering and it didn't complain at me. I could send you the two FLAC's if you wish already unzipped.

 

Thanks for helping others in D/L procedure and the test as well.

 

You can PM me with your vote, tell me how you would have like to vote, and I will keep it in mind for the final counting

[bibo01]

In case you have problems to download the two tracks from the original link, you can download them also from here:

https://www.dropbox.com/s/i4pb51k3r6etyr6/test%20ascolto.zip?dl=0

[bibo01]

To be precise for everybody:

the two converters are the same, with the same analog output stage, only the clock of one of them is damaged and generates more jitter.

[mmerrill99]

To be precise for everybody:

the two converters are the same, with the same analog output stage, only the clock of one of them is damaged and generates more jitter.

Thanks for that explanation - it makes much more sense of the test as I was contemplating that the test was just two different DACs i.e everything different & this would negate the validity of the test in many people's mind & probably prevent many people from engaging in it.

 

I think it would be wise to highlight what you have just posted in the different forums you have posted the poll

 

From the graphs the base of the main signal has a much wider spread than the other - which signifies significant close-in jitter before consideration is even given to the side spurs.

 

The "normal" jitter test is the Jtest - where a main signal tone (usually 24KHz) is combined with another signal which toggles the low-order bit (LSB). This was originally designed to test the data induced jitter inherent in SPDIF receivers but is still used as a rough jitter test where no SPDIF receivers are involved i.e. USB receivers in computer audio

[bluesman]

Out on a limb but committed, I can definitely hear a difference between the two tracks. I wish I knew what the instruments sound like live, though. I think the flute on A1 is clearer and a little bit "thinner". Attacks are cleaner and notes are a bit better defined. A2 sounds a bit muddy & fat by comparison, and is not quite so clearly a flute until a few notes have come out of it. But the tone of the flute A2 does sound more "sinusoidal" to me - so does it have less odd order harmonic content in its native tone? If so, then the jitter is in A1 and it's adding odd order harmonic distortion. But through my earbuds at work, A1 sounds more like a flute than A2.

 

The harpsichord is closer on the two than the flute, but it sounds like a better instrument on A1 than on A2 - the low register is a little richer and the pluck a bit cleaner. So I'm sticking with A2 as the jittery one.