Dave, Maybe! Warning: I may not know what I'm talking about here. I know just enough to be totally dangerous, and I'm writing from memory. Corrections encouraged. CD playback is a complex electromechanical operation that cannot be totally free of jitter. Of course, there are many things that are done in modern CD units to reduce jitter, but it cannot be totally eliminated - it's the nature of the beast. The laser is tracking the "pits" (actually 1/2 wavelength high bumps), and using these to figure out the linear speed the pits should be running at, then correcting the rotational speed in proportion to the current diameter of the pit track. Correction means servos in a feedback loop. The speed is always being corrected. There's an old saying about closed loop engines, and that is that they are never running at the right speed. I think adding weight to the rotating disc is damping the fluctuations in speed. But of course the process is a lot smarter than that. The bits are fed into a buffer in the DAC then doled out with the timing set by the world clock. It's a lot more accurate than deriving the clock from the bit stream, but it does have jitter. The latest crystal oscillators run at much higher freqencies more accurately, so when used to drive the clock through a ratio end up with lower jitter than was possible 10 years ago. If the DAC's world clock is used to assist the laser mechanism that reads the bumps and sets the rotational speed, jitter is further controlled. Regards, Metralla
If you rip a cd into your computer, is the jitter "copied" (for lack of a better word) into the bits that are now on your computer? Or does everything come in perfectly? If perfect, then when you play the wav (or aiff on the mac) files does that mean there's no jitter? HZ
The PCM standard utlized in commercial CD audio reproduction is beautiful in its execution. Perfect it is not, but the digital format utilizes complex but hightly effective strategies for dealing with time and space. Those who put the pieces together for this technology read comments about timing errors related to jitter and must roll on the floor holding their sides. "That's a good one". Time to pull out my red marker to tame that jitter (bug).
Jitter is complicated and complex, and I don't pretend to understand it. But I think it's primarily a playback issue, when the digital signal is converted to analog. I don't think it's normally a case of jitter being in a CD itself, and I don't think any jitter is introduced by extraction.
Ok.. if that's the case, then playing a ripped cd that's stored in memory wouldn't have jitter, correct?
If that's the case then why are all kinds of CD burner manufacturers and manufacturers of CD transports addressing this problem? Because of a few nuts like me? A few years ago I heard all kinds of fellows like yourself saying it doesn't exsist at all. So why are manufacturers attempting to correct the nonexistant? Like it or not, you directly benefit from the guys with the markers in their hands.
That would seem logical wouldn't it? If you look at many of CD burning solutions they all offer some sort of jitter correction. What sucks is that jitter can be added through a lot of different ways into the playback media. Including introducing it at the time of the CD burn, as Plextor and Yamaha engineers in particular believe. Sometimes, the design of the playback electronics themselves can introduce jitter, even if you've done your best to eliminate it at every other part of the chain. Some units are supposed to hopelessly be flawed to always produce jitter because of the PPL devices in their signal path. So what to do? You can pretend it doesn't exist... LA LA LA LA LA LA LA LA LA (I can't hear you!) LA LA LA LA! Do as Humorem sez-- buy a turntable... or you can choose kookdom... or is that kook-dumb?
Seems like there's a third way out: have a custom computer setup that sucks the cd in and then plays it from memory instead of the original cd. No jitter, right? Since ram is so cheap nowdays, $200 should get ya a gig, and a cd only holds what, 650 meg? You can suck the whole thing into ram.
I posted these links on the DCC forum. Here are they are again: http://www.geocities.com/jonrisch/jitter.htm http://www.angelfire.com/music4/audio_visions/jitter.html
You're on the right track, I think, but then some other jitter will get in the way! Yes, when you rip the contents of a cd into a wave file (for example), you leave the original medium's jitter behind. It's gone. All you have is the data itself. And that's good, right? But remember, the 'jitter' we are talking about is an audible problem only during the d/a conversion of the data. So when you play that data off your hard disk, your computer introduces a different jitter characteristic than when the same data is read from the cd itself. Is that "better jitter" than the original jitter from the cd? Maybe yes, maybe no, but it's certainly different. Likewise, when you burn a cd from that data on the hard disk, your system introduces a different jitter characteristic into the pits on the cdr. This is one big reason why a copy will not seem to sound exactly the same as the original. If you are lucky, it might sound a little better than the original! --- What's really annoying about jitter, after all these years, is that its effects have not been completely stamped out through better engineering. You would think that by now, only the cheapest crappy digital equipment would have any audible problems with jitter.
I like that first one. But the 2nd one loses me when he insists that DAT-to-DAT dubs have no effect on jitter.
I have not hooked the Mac up to the stereo to check this (do not trust headphones thru the front of the machine as the motherboard bringing the signal up to listenable level sounds a bit thin), but I have read several places that jitter is a problem playing aiff / wav files from a hd. Playing mp3's is another matter entirely; the playback software has a "sound" of it's own and that is all the distortion you should hear if it is a hi-bitrate file. Extracting a cd always forces your burner / burner software to reclock the data. This easily explains how the dub can be a slight improvement on the original, in this format. FabFour is right that the jitter gets into the waveform at the d/a conversion stage. Here's a funny: I have the lp The Saints "All Fool's Day" on TVT Records. I also have the two tone black cover version bootleg lp of Prince's "Black Album". These lp's sound like cd's played on an early cd player, like a Sony CDP-101, because that is exactly what they were each mastered from. No one has to tell me this, they both sound like jittery cd playback, right thru the vinyl.
I think that would be a function of the software, and whether you have slow/old hardware or not. I mean, the software could always suck the entire song into memory and throw it at the d/a converter. As long as your cpu is busy doing something else (say, reading the SH Forums!) you'd no longer have physical rotational based time-jitter problems, right? Again, I really don't know anything about this topic - it just *seems* that one could eliminate the problem by cacheing the bits so that they are always instantly available.
Don't quote me, but I believe that at least to a certain extant jitter is a problem with the timing of the D/A (or A/D) conversion. For example, maybe the clock on the converter should advance every 1 ms. Well, sometimes it advances .1 ms too fast and sometimes it's .1 ms too slow. It's hard to describe without drawing a picture, but what that means is the waveform isn't being properly reproduced. This isn't necessarily a problem with the reading in of the data, but rather the accuracy of the clock on the D/A converter... Anyone?
"The PCM standard utlized in commercial CD audio reproduction is beautiful in its execution." Damn! That's what I was reading twenty years ago! Whither progress? "...the digital format utilizes complex but hightly effective strategies for dealing with time and space" Does anyone have proof of that? I'd settle for peer-reviewed double-blind tests. "Those who put the pieces together for this technology read comments about timing errors related to jitter and must roll on the floor holding their sides." I actually know some people who *were* involved in the effort and indeed they are laughing -- at Sony/Philips for stupidity and cost-cutting strategies! clark
Heck, no one suggested that the CD standard or its execution is perfect; whatever that means. But Sony and Philips were not stupid. Sony insisted in expanding the Philips proposal to 16 bits and the size to accomodate longer classical pieces. Some of those old guys involved in its development have disclosed in interviews over the years that the CD standard met all the criteria placed before them and the criteria were selected to meet the sound quality standards thought to be relivent. They also noted that if others hear things they do not they can accept that. I believe Audio Mag carried several good interviews on this over the years. I seem to prefer the sound of several of my SACD's so there just may be something to those who claim to hear something going on that is not an issue with the SACD format. Be it timing errors caused by some form of jitter, analog to digital errors not related to jitter, etc. I don't think anyone knows why and it may be some time until there is even a majority opinion. But then, I also like the sound I hear from many of my well produced and mastered conventional CD's.
That's correct Luke! In fact, several years ago, Sony/Phillips thought jitter was a serious enough issue that they proposed a modification to the red-book standard for it. I don't know if that actually happened of if it was just considered. Anyone?
Don't know if its at all jitter related, but the first generation and even second generation CD players had sound quality issues. One of the Sony developments with its 201 second generation unit was a change in the timing clock to sync both channels together. It's been a long time so I don't recall Sony's hype it issued concerning this issue, but that and a few other changes sure made the units sound a lot better to may.
Just curious, how many people here are using older (10 yrs or older) CD players as opposed to newer models.
Whether you read the bits from a CD or from memory you still have the problem of sending the samples to the d/a converter every 22.68 microseconds (ie 44.1 thousand samples per second). In fact, when playing a CD, the bits are stored in a buffer, before being sent to the d/a converter so you could think of playing a CD as the same as playing the data directly from memory! It is the difficulty in designing a sytem that can time the samples sent to the converter as accurately as possible that is the problem and this is just as much an issue in a sound card as it is in a CD player.
The Black Album bootleg? Thomoz, About that Black Album bootleg, besides it sounding like 'jittery cd playback' as you described, how was the overall sound quality? I remember having a bootleg of that on cassette that the owner of the little indie record shop I used to frequent taped; on the other side was a bootleg of a live show. It sounded like garbage to me, but that could have been because of the cassette deck used to dub it and/or the original source. It was dull, noisy and totally lacking in dynamics. It didn't sound like it was dubbed from vinyl, nor did it sound like it was dubbed from CD. Now, the official release of the Black Album was an eye opener; not that it sounded great or anything, but compared to the bootleg I had, it was like night and day.
"Don't know if its at all jitter related, but the first generation and even second generation CD players had sound quality issues." Understatement of the year! But if I recall correctly, my screaming my fool head off about this was met at the time with stony silence from the AES, the Major Manufacturers, the Compact Disc Group, the Major Media, Audio and Stereo Review, the Major Retailers, the Boston Audio Society... Oh it was terrible. Apparently not a one of 'em could listen! Or was it... Satan!? clark
I guess no one wanted to respond to my question. The reason I asked is because many audiophiles' only impression of CD comes from those early players. Some are still using old players.
Luke: you nailed it, that is exactly the story of jitter's effects on d/a playback. Grant: I have 1 9 y.o. cd player (4x oversampling, after a p.e.m. filter to square up the waveform), a 2 y.o. dvd player (Toshiba), 2 5 y.o. cd roms, a 4 y.o. cd rom, a 2 y.o. portable (Panasonic 1-bit), a Yamaha cd-r player / burner, two 2 y.o. boomboxes, a brand new desk top stereo mini-system, and a 2 y.o. Pioneer MOSFET head unit in my car. The inked cd's show marked improvement in all the players except the dvd, and that one supposedly upsamples to 96khz before d/a conversion. Nevertheless, the bass/treble ends are pronounced and the upper-mids very recessed on this player so it is unsuitable for playback of all but a dozen of my music cd's.
