[nick_charles]

Quote:

Originally Posted by regal /img/forum/go_quote.gif Not only that but in the real world Jitter isn't concentrated at a single frequency it is spread throughout the audio band making it even more impossible to hear.

Yes and no. If you look at the tests that Stereophile has done on CD players there is a wide variation in patterns, some have very evenly spread jitter that is essentially low level background noise, others have a few strong sidebands around the fundamental and then at harmonics with decreasing intensity, some have a mix of both. The pattern for the oppo is the worst I have seen.

Quote:

A computer harddrive is superior to a CDP because with EAC there is no error correction being done as we play our disks, error correction is audible.

EAC also does error correction but it is just different, EAC homes in on C1 and C2 errors and can reread sectors multiple times to get a best guess but it is still possible to have errors get through if your disc is mangled enough, they are rarely audible though.

CDP error correction is inferior in that the disc has to be read and rendered in real time, but the instance of uncorrected audible CD errors is pretty low, I do not speculate on how audible the correction process is but I rarely hear audible blips these days.

As for just having the wrong bit value, this would only be an issue if you had a whole string of them one after another and if the errors were predominantly in the high order bits. For instance if the signal is

01000000001111 and the cd player sees 0100000000001110 that is pretty negligible but if it was 0100000000000001 and the CD sees 000000000000001 well that is a big issue but that is 1/44,100 of a second so you would need a lomg stream of these bad values to have an effect ?

How would you know the difference between error correction and just the different sound of different CD players. I do not know of any research that has looked at error correction and listeners ability to detect it but if you have some sources I would be interested to read about it.

I remember that some of the early CD players had little error correction LEDs that flashed to let you know the error correction was working.

 

[bigshot]

Quote:

Originally Posted by regal /img/forum/go_quote.gif A computer harddrive is superior to a CDP because with EAC there is no error correction being done as we play our disks, error correction is audible.

C1 error correction is totally inaudible, and that accounts for the lion's share of error correction. C2 error correction is quite rare for relatively clean disks, and it would only be an instantaneous thing. It wouldn't account for people's reports of jitter causing "reduced soundstage" and "veils". Those sorts of things are caused by people listening too hard. They start thinking they hear things they don't.

See ya
Steve

 

[OblivionLord]

This is a very long thread and alot to read... can someone link a site that explains jitter so I dont have to waste anyones time? Thanks

 

[bigshot]

Nick Charles' last few posts have pretty much nailed it.

See ya
Steve

 

[tfarney]

Well, I made it through 24 pages and came to the only logical conclusion: Jitter is the new cable. Not that it isn't real, but the jitter proponents don't serve themselves well with the same old positioning: Summary dismissal of data that contradicts their view, regardless of the quality of the source, descriptions of what they hear with the same old elusive terminology (amounting to attributes that could belong to almost anything) and dismissal of all disagreement with condescension: "Oh you can't hear it? Isn't that nice for you? You'll save a lot of money, won't you?"

In the final analysis, even the low-rent systems found in high-end audio salons are not resolving enough to hear what they hear. So only a very small group, consisting of the most obsessive audiophile hobbyists and their vendors, are in this exclusive club, and we, the ignorant unwashed, are left to decide for ourselves who is hearing what they want to hear -- those who need to justify the expenditure of thousands of dollars (and those who sold it all to them) in the face of the possibility that they may have spent it on phantoms, or millions of music lovers and quite a few studio pros. It's not hard; not for me anyway. The jitter proponents may even be right, but they are not credible.

Tim

 

[bigshot]

Get ready for trouble tfarney. Common sense doesn't sit well with some folks.

See ya
Steve

 

[tfarney]

Well,I read the rest of it (slow week), and I've come to an even more useful conclusion:

On one side we have the "jitter is not an issue" contingent, headed by Big Shot and the late-arriving, but informative Nick Charles. These guys have offered a fair amount of data and referenced what appears to be pretty scientific and well-verified study. And of course, that doesn't make them right.

On the other side, we have the "jitter is audible and distorting" contingent, headed by Audioengnr, who questions the validity of the study, has offered much anecdotal evidence, but has, thus far, failed to counter the "flawed study" with any similarly scientific data that refutes it. And of course, that doesn't make him wrong.

But if you read the whole thread and paid close enough attention, you will have noticed that in refuting Big Shot's study, Audioengnr said they probably weren't using any systems resolving enough to reveal the jitter, they were probably only using, at best, the kind of stuff found in the demo systems of your typical audio salon. Not Best Buy, or even the good HT store with a couple of nice two channel systems in the back room. According to the leader of the jitter is an issue group, hearing it enough to validate the study would require something better than the demo system in the average audio salon.

So we have two sides to this argument:

1) Jitter levels, even on just average consumer equipment, are below the level of hearing.

2) Jitter only becomes audible on high-end systems high enough to be beyond the capabilities of the systems used to sell high-end equipment in most audio salons.

In a twisted sort of way, they agree. And at least we can agree that it doesn't matter to most of us. I mean, how many of you have the money to aspire to something beyond the demo system at your neighborhood audio salon?

Exactly.

This one got to 37 pages over something that will never be an issue for the overwhelming majority of us. And if the guy who owns a system more expensive than a high-end shop's demo system can afford to drop another grand or two on a re-clocker, it really doesn't matter if it reduces jitter audibly or simply makes them feel better to know it has been reduced. This is what they do.

Tim

 

[spraggih]

Quote:

Originally Posted by tfarney /img/forum/go_quote.gif Well,I read the rest of it (slow week), and I've come to an even more useful conclusion: On one side we have ... On the other side, we have ... But if you read the whole thread and paid close enough attention, you will have noticed that in refuting Big Shot's study, Audioengnr said ... So we have two sides to this argument: 1) Jitter levels, even on just average consumer equipment, are below the level of hearing. 2) Jitter only becomes audible on high-end systems high enough to be beyond the capabilities of the systems used to sell high-end equipment in most audio salons. In a twisted sort of way, they agree. And at least we can agree that it doesn't matter to most of us. I mean, how many of you have the money to aspire to something beyond the demo system at your neighborhood audio salon? Exactly. This one got to 37 pages over something that will never be an issue for the overwhelming majority of us. And if the guy who owns a system more expensive than a high-end shop's demo system can afford to drop another grand or two on a re-clocker, it really doesn't matter if it reduces jitter audibly or simply makes them feel better to know it has been reduced. This is what they do. Tim

A great read and study of the argument IMO. I hope you do litigation, or work at the UN or are a dad or work in some field where your skills are needed. Your ability to see one person's point of view in isolation and then to look at another point of view in isolation and find agreement, a way forward is very good.

Not to say others do not have the same skill and concluded this on post #289 or post #112 but nonetheless this does not take away from your skill.

Best!

 

[bigshot]

Quote:

Originally Posted by tfarney /img/forum/go_quote.gif But if you read the whole thread and paid close enough attention, you will have noticed that in refuting Big Shot's study, Audioengnr said they probably weren't using any systems resolving enough to reveal the jitter, they were probably only using, at best, the kind of stuff found in the demo systems of your typical audio salon.

If he had bothered to pay the five bucks and get the pdf file, he would have found that the testing took place at the homes and studios of audiophiles and professional audio engineers using their own carefully tweaked reference rigs and listening rooms. The testing took place on the most "resolving" equipment available. He was arguing against the paper without having ever read it. What does that tell you about his argument?

In my experience, the truth rarely lies halfway between two opposing viewpoints.

See ya
Steve

 

[tfarney]

Quote:

Originally Posted by bigshot /img/forum/go_quote.gif If he had bothered to pay the five bucks and get the pdf file, he would have found that the testing took place at the homes and studios of audiophiles and professional audio engineers using their own carefully tweaked reference rigs and listening rooms. The testing took place on the most "resolving" equipment available. He was arguing against the paper without having ever read it. What does that tell you about his argument? In my experience, the truth rarely lies halfway between two opposing viewpoints. See ya Steve

Yes, but I was neither looking for the halfway point nor the truth. Just resolution. Even if he read the study and found another that refuted it; even if he presented irrefutable evidence that jitter is audible if only every component in your system is resolving enough, the standard is too high. Probably too high for nearly every participant in this thread, even the ones who hear jitter, and who, I'm sure, missed that part of his argument...

"I hear jitter, people..."

...and with a single image, i have blown my reputation as a peacemaker.

Tim

 

[bigshot]

The halfway comment was intended for spraggih. Sorry for the imprecision in my replying technique.

See ya
Steve

 

[b0dhi]

Quote:

Originally Posted by nick_charles /img/forum/go_quote.gif But are even 80db down jitter sidebands audible ?, the evidence so far would suggest probably not, I found this site which hosts some samples of sounds with jitter added at different levels. PCABX 60 Hz Jitter Demonstration I woud suggest that those interested download the unjittered and -80db jittered signals and try seeing if they can detect the jitter , in a blind test of course, personally I could not, and that is much much worse (32x worse) than anything you are likely to get in the real world...

I couldn't access that site, but judging by the title, it's based on jitter in a 60Hz wave. Are you aware that distortion audibility is orders of magnitude poorer for bass tones than, for example, a 4Khz tone? Not only that, but for a 60Hz wave, it's time-domain resolution is predominantly affected by quantisation precision (to within picoseconds), not sample precision, so jitter would have less of an effect for this waveform in any case.

A better subjective test for jitter audibility would be a square/saw wave (ideally with as much bandwidth as possible) around 10Khz or so.

 

[nick_charles]

Quote:

Originally Posted by b0dhi /img/forum/go_quote.gif I couldn't access that site, but judging by the title, it's based on jitter in a 60Hz wave. Are you aware that distortion audibility is orders of magnitude poorer for bass tones than, for example, a 4Khz tone? Not only that, but for a 60Hz wave, it's time-domain resolution is predominantly affected by quantisation precision (to within picoseconds), not sample precision, so jitter would have less of an effect for this waveform in any case. A better subjective test for jitter audibility would be a square/saw wave (ideally with as much bandwidth as possible) around 10Khz or so.

The site seems unavailable now, it was there a few days back, probably just maintenance, I do have the files myself and if the site does not come back up I will host them.

No, the signals are music signals, two samples one solo piano and one castenet, the jitter is a 60hz signal.

I am aware that jitter audibility is frequency dependent, Dunn's model puts the most susceptible frequencies at 20K , 20ps by his model, however most tests that use a pure tone instead of music use an 11Khz signal as the fundamental. Dunn's model however depends on a fundamental signal that is 120db above the threshold of hearing, so that theoretically 10ps would be audible at 20Khz at a listening room volume of about 180db for someone with superb hearing. The empirical figures, however do not support this model and Dunn acknowledges this in his later papers.

A better subjective test than a square wave that would be music, which is what both the Dolby Labs and JAS studies used, the Dolby labs study used both pure tones and music samples. With pure high frequency tones jitter at 10ns was audible, in music it had to be 20 - 30ns, this was deterministic jitter, with random jitter as in the JAS studies the thresholds are as you would expect much higher as the effect is evenly spread and equivalent to low level noise. Now if you want to pick nits you could argue that this random jitter would be more audible as noise if you cranked up the volume to, and I mean this literally, ear-bleeding volumes.

For reference here is the jitter performance from a few random CD devices including the venerable Sony CDP101

PC AV Tech Digital Player Report: Sony CDP-101 CD Player

PC AV Tech Digital Player Report: Sony D-220 Portable CD Player

PC AV Tech Digital Player Report: Panasonic SL-S220 Portable CD Player

Marantz CD-67SE CD Player Performance

PC AV Tech Digital Player Report: Sony CDP XE 500 CD Player

PC AV Tech Digital Player Report: Sony D-828 Portable CD Player

PC AV Tech Digital Player Report: Teac PD-75 CD Player

I deliberately chose the worst possible figures for each. This is a collection of good bad and indifferent figures, some **are** pretty bad, however none are what you would call modern machines and the earliest one the Sony CDP101 actually does very well. For more modern comparators Stereophile does comprehensive technical testing and has charts for jitter, below I have included a link to jitter measurements for the legendary Radio Shack PCDP and a decent but hardly esoteric modern machine (Onkyo DX7555) and the notorious Oppo...

Stereophile: Radio Shack Optimus CD-3400 portable CD player

Stereophile: Onkyo DX-7555 CD player

Stereophile: Oppo DV-970HD universal player

 

[tfarney]

Actually, the best test would be an objective one -- a double blind listening test of music being played on a reference system, with and without devices like audioengnr's re-clocker, that have (I assume it has...) been measured significantly reducing jitter. I know DBT is a dirty word in the audiophile world, but if you run the tests up to a statistically significant sample and the listeners can't correctly identify the presence of the distortion more than 50% of the time, it's not audible. Simple as that. If someone thinks detecting distortion requires "ear-training," well, ok, use audiophiles as listeners. If they think the distortion only reveals itself through long, careful listening over time and, therefore, the inability to hear the distortion doesn't mean it's inaudible, then let them spend their money as they will.

We're not talking about a subjective art here. We're talking about the reduction of something that is measurable and the claim that the reduction is audible by humans. That claim can be tested, and either verified or proven false.

Tim

 

[Pars]

I thought you guys had summarily dispensed of jitter as even being a factor worth investigating?

Also I don't see any jitter measurements of transport/DAC combos in those posted. Those could be interesting.