Jitter!!! please let me listen some!!!

[Level10Pro]

Hi guyz, may I chime in?

Hey, I really love the fact that we can actually talk real engineering for a change. I am/was schooled and trained in digital engineering and I have dealt with issues such as jitter half of my adult life. So on that happy note, I have to agree 100% with what Monomer is saying! Some of those cute little animations that Space has here do show what jitter kind of looks like.

Space, if what you were seeing was jitter, then each cycle of your test would display slightly different A/D steps from cycle to cycle. I'm not seeing that. Just differences in sample test results. I think you would have to maybe put an O-scope on the analog output of your DAC on a very high frequency test sample and look for actual 'jitterring' to occur, and when you start talking about 1 part/million errors, man, you better have some pretty good gear! And whether or not you can hear it becomes more and more subject too.

I guess what we really care about tho is more about the A/D though, so unless you have access to be able to put a scope directly on your clock, isn't it kind of hard to actually measure true jitter anyway? I don't even see any signs of jitter on Space's test results at all. Just different sampling anomalies that are more dependent on filters and circuitry, timing between A/D and DAW and who knows what, but Jitter???
I'm only refering to the DAW images, not the scope ones, as I'm not even sure what it is we are looking at.

[space2012]

Originally Posted by Level10Pro

when you start talking about 1 part/million errors, man, you better have some pretty good gear! And whether or not you can hear it becomes more and more subject too.

AES.Org AES11-2003 paper say AES Grade1 Clock is 1ppm,
any TCXO or OCXO Xtal has <1ppm at 31°C with clean & stable power.
Rubidium Atomic Clocks have parts per billion +-0.03ppb.
acording to Teac Esoteric and Antelope 10M specs.
....
tested 2x times the M-Clock, and the diference between M-Clock tests were less than the RME Clock diference.
you can see in the screen shots:
M-Clock +4dBu,
M-Clock -16dBu.
two diferent tests with same clock.

The M-Clock was not calibrated in those tests, had drift.
recently did a re-calibration to factory specs.
YouTube: WordClock Drift Test

[akai96]

Originally Posted by living sounds

Just listened to those examples, the difference is very obvious. The RME in this example is a far better converter. Attacks sounds a lot more blurred on the Motu example. However, this might or might not be mostly because of the clock. In general it is consistent with my experiences with clocking.

It is the clock,because I tried the lucid on the motu 828.thumbsup

[Chip Stylus]

I've heard jitter cut into vinyl on two occasions: these albums were sourced from digital
and it sounds like they used a cd, played back poorly, as their "tape source"

The albums are
Prince - Black Album (late eighties bootleg) 1988?
Saints - All Fool's Day (TVT Records) 1986?

The Saints record is not hard to find, and when you play it
(you DO still have a vinyl rig, don't you?) it's really obvious that
the lp sounds like a badly rendered cd - with surface noise .

[just.sounds]

Originally Posted by monomer

Jitter is not constant, it is random.

You act as if jitter is the only possible flaw in digital audio.
It is not.

And please stop posting more screenshots and links.

OR it is audio source related wich is the worst kind.
But it is clearly audible when you have the possibility to A/B

Classic stereo recordings (AB) suffer the most from jitter.

Sound differences:

Narrow stereo field
haze
and this is more subjective but it doesn't feel as tightly timed musically speaking.

[rolo]

I have a DIGI HD3 Accel system with Genex and Apogee converters, as well as some External effects processors. I always clocked everything with the MAC's/Digi's internal clock. Last month I bought a Lucid 192 Clock Distro.The depth and imaging difference was not subtle. Everyone in the room noticed a big change for the better.
cheers

[Tom Hakala]

Originally Posted by rolo

Everyone in the room noticed a big change for the better.

how's your room? and it didn't even matter where in the room everyone sat? And who are this "everything".

[rolo]

Originally Posted by Tom Hakala

how's your room? and it didn't even matter where in the room everyone sat? And who are this "everything".

my room's fine thanks, just getting over a cold, but doing much better. everyone sat on my lap. "who" and "everything" are folded neatly in my front pocket.
listening in were the st louis barber shop quartet and their mothers. all wearing plaid.

would you like to know what color undies i was wearing?

[Tom Hakala]

just asking. Because words "everyone" and "big" caught my eye. But I think if the same group would do a blind test... then it wouldn't be that big. Did you do blind testing? And did you listen to them back to back?

[rolo]

yes we did and indeed EVERYONE noticed a BIG difference. you don't have to believe me. this is just a forum with opinions. the OP asked for others personal experience in the matter... this was mine and others here.

[rolo]

Originally Posted by Tom Hakala

I think if the same group would do a blind test... the it wouldn't be big.

your post is merely speculation. perhaps you should do a test with the proper gear, then state the facts.

[Tom Hakala]

Originally Posted by rolo

your post is merely speculation. perhaps you should do a test with the proper gear, then state the facts.

I'm not stating anything. I'm questioning what you said.

Did you listen back to back?

[rolo]

yup.. clock and non clocked.

[Tom Hakala]

Originally Posted by rolo

yup.. clock and non clocked.

and it was a blind test?

[rolo]

are you kidding me? that is just a silly question. tell ya what.. c'mon over put a blindfold on and you be the judge. cookies and milk for everyone!

[Tom Hakala]

Originally Posted by rolo

are you kidding me? that is just a silly question. tell ya what.. c'mon over put a blindfold on and you be the judge. cookies and milk for everyone!

so if it was a silly question, I will assume that you did do a blind test. If the audible difference is that big it can't be jitter. There is differences, but how audible those are is another question. And punch of people in a room, all hearing it CLEARLY... and saying it's BIG...

[rolo]

i just love how you declare your opinion with no personal experience to back it up. you were not here, you dont own the studio i own, you don't have the gear that i have, and as far as i can tell, you have never done a test like this at all. it's opinions like yours that guide people looking for experienced knowledge in the wrong direction. i've read some of your past posts. you don't use high end pre amps because, according to you, they make no difference. it's clear what i'm up against. i've stated my personal experienced opinion. whether you agree or disagree is your deal, but until you've tried it out personally, well... you know what you can do.

[Tom Hakala]

Preamps. I don't like the idea that newbies are guided to buy high end preamps. Just go to the low end forum and you will see. I have had high end preamps, I have access to high end preamps. But it's never the preamp that I will go after if the sound isn't right. There is differences between preamps, I have heard differences and the same differences can be measured. I have blind tested A LOT. But the differences are not night and day (the more distortion the clearer the difference). My opinions about preamps aren't relative. Because..

the subject here is jitter.

I'm must be clearer now.. and rethink.. ok, I can't say you guys didn't hear a difference. I can't know that, and there's nothing that can prove it wrong. BUT what I really doubt is that the BIG difference was jitter.
I also could say that I have tested different clocks with people, and we couldn't hear a difference. But it wouldn't get us anywhere.

So can you somehow prove your point, and prove that my doubt is wrong that jitter is really that audible. Because that would end the long running argument between is there or isn't there AUDIBLE difference... Saying that "me and my friends all heard that so you really have to trust me" won't do it.

[rolo]

brotha... i really don't have to prove jack to you at all. i find your opinions lacking in experience and common sense. i'll say it again... FROM EXPERIENCE... I, along with other people... Producer, 2nd engineer, and the artist noticed a big difference in depth and imaging.
furthermore, i take back my invitation to my studio so that you can do your BLIND tests you are so fond of. No cookies and milk for you.
btw... just out of curiousity,is there any place i can hear some of your recordings?

[Tom Hakala]

I'm fairly new to this whole recording thing. You must be someone who is doing a nice living out of this. But it's still besides the point.

You say my opinions are lacking common sense (well it might, mostly because my English I guess)... but you still are knocking blind tests. If you take a look at the debate that's going on here on GS or different forums, there's really no AUDIBLE proof about jitter. But you and your colleges can easily hear it, and the difference is big. But you are not willing to prove it nor do a double blind test. To me it just doesn't make any sense, no matter if I'm not a super pro...

And remember, I'm not saying that you guys weren't hearing a difference. But the difference most likely wasn't about the jitter.

[badboymusic]

Originally Posted by HookedOnHardware

YouTube - illusion of superiority

Thanks, I'm saving this one!

[space2012]

Originally Posted by Tom Hakala

there's really no AUDIBLE proof about jitter. But you and your colleges can easily hear it, and the difference is big. But you are not willing to prove it nor do a double blind test. To me it just doesn't make any sense, no matter if I'm not a super pro...

And remember, I'm not saying that you guys weren't hearing a difference. But the difference most likely wasn't about the jitter.

what does not make sense is that you cant hear it.
what does not make sense is that you say "the difference most likely wasn't about the jitter." but cant hear jitter.

test A/B
true pure analog vs. -->AD-->DA-->

with a condenser mic
an small 100% analog mixer, behringer, mackie, alesis.

connect the mixer to the loudspeakers.
connect the condenser microphone, "any rode nt1, adk, mxl v63m will do"
avoid feedback by placing the mic in another room.
ask someone to talk, read or sing.
send the analog signal aux.send to the AD converter.
activate software monitor input in Cubase/Nuendo, Sonar, Reaper, Samplitude, Traktion, SAW, etc...
set buffer size as small as possible, usually 128 samples.
connect the DA converter to the console
and test A/B, Solo/Mute Pure Analog vs. Digital.

but if you cannot hear the difference:
Coup de grâce - Wikipedia, the free encyclopedia
Seppuku - Wikipedia, the free encyclopedia

[Tom Hakala]

Originally Posted by space2012

what does not make sense is that you cant hear it.
what does not make sense is that you say "the difference most likely wasn't about the jitter." but cant hear jitter.

I'll ask you the same question as Ethan did here:
http://www.gearslutz.com/board/4487667-post294.html

[psykostx]

Jitter isn't a specific thing to be heard and picked out, it effects everthing from system stability (dropouts) to frequency balance (as you can see from the graphs on page 1 of this post, errors in transients can add or subtract significant amounts of high frequency energy). Jitter can be heard as overall sound quality. Think of a streaming MP3 on a bad internet radio station. It sounds fuzzy, scratchy, sometimes too boomy, sometimes too thin (they all sound that way on my system). That is the sound of "aliasing" a possible side effect of jitter. I say "possible" because jitter can effect everything and everything depending on the resolution of the rest of your gear, or it can do almost nothing to change the perceived audio signal (if your rig is distorting your signal even more before it hits your ears). Needless to say, jitter is a measurable real thing, though it may not always be audible, or its effects on perception predictable. One poster said that jitter ruins the stereo image. This is true provided that you have a highly phase accurate, low distortion, monitoring system. Basically, jitter is like noise or harmonic distortion, it can actually sound pleasing, even "better", but it gets in the way of the truth. You won't notice jitter outright unless you have quite a high resolution monitoring system... but everyone will notice jitter in their final mixes if it exists in their system. Its the thing that makes your mixdowns sound like MP3's before they become MP3's...it really can ruin percieved volume, depth, hell, percieved anything. That said, don't use optical unless you have a really good transmitter and receiver. Use AES/EBU balanced XLR to transfer digital signals. It really eliminates most of the jitter, no matter what clock or converter you are using.

[Ben B]

The only "science" I've ever come across on this subject clearly states that jitter manifests as FM sidebands in the signal, some number of dB below a given input signal, depending on the jitter amplitude. The amount of attenuation is relatively easy to calculate.

If sidebands are the only artifacts that manifest in the signal (and I haven't seen any hard proof to the contrary), you could get a decent idea of what jitter sounds like by using a ring modulator.

Duplicate a track in your DAW (single instrument or full mix), and run a ring modulator plug-in on the duplicate. Create the most obnoxious sidebands you possibly can. (Feel free to modulate them as well. Create a worst-case scenario.)

Now blend this distorted signal under the first undistorted track at varying amounts of attenuation. This effectively simulates what changes at varying jitter amplitudes. By doing this, you'll get a pretty good idea of your own personal detection thresholds for the type of distortion jitter would cause.

Suppose you do this experiment, and find that after the distorted duplicate is attenuated by 60 dB, you can no longer hear its influence on the main signal. That would be your personal detection threshold for that particular mix. You could repeat the process with different kinds of signals, with different frequency spectra and dynamics, if you think the distortion would be less likely to mask under certain conditions than others. Find out what your minimum detection threshold is.

You can then compare your detection thresholds to the amount of actual sideband attenuation that is present in real converters as a result of clock jitter. You will probably find that the real-world attenuation is greater (in many cases, far greater) than your own personal detection thresholds, in which case the differences you're hearing among different converters are more likely to be caused by other factors (e.g. variations in quality of the analog components in your particular sound card or converter).

If anyone has any non-anecdotal references to articles which establish (scientifically) that there are artifacts caused by jitter other than FM sidebands, please post them here. By non-anecdotal, I mean that I'm not looking for "converter shootouts" or listening tests, etc. I think these lead to more confusion. For example, when people listen to two converters and hear a genuine difference, they often prematurely jump to the conclusion that the sonic differences are due to jitter (when they are more likely caused by other factors).

So, if there is other science to this, please share what you know!

-Ben B