CD ripping with jitter correction

"In the transport/DAC world, jitter can result in distortion."

What's the point? this is what I have been saying here and elsewhere.

"After that, the only 'jitter' one should be concerned about is jitter on the burning side."

For the most part, this is what I said. The thread stated that burning was also involved which makes Katz articles relevent. The only thing I will add is that after transport DAC, you can also have cable jitter and pressing induced jitter. These also have sonic effects.

""n+1 copying effects relate only to realtime copying, not rip and burn."

I disagree, n+1 copying also affects burning as well. I have heard it myself on several occassions.

See the Chesky test CDs for an example of this. Also, jitter in the process of manufacturing of discs is a big issue. A recent email from Michael Bishop confirms the problems they have had in disc mfr and that they were solved by Sony Disc Manufacturing...

 

Please explain how n+1 effects can effect burning when the data file containing the music does not have timing data, only musical content. The clock of the buffer in the CDR drive, and the accuracy of the mechanism itself can create jitter, but that jitter has nothing to do with the source, so there is no n+1 effect.

 

Nothing weird about it, Grant--it's basically a free product, constantly an unfinished "work in progress" and, in my own experience, it shows in the form of software bugs, poorly documented features, or rips that come out with glitches. This behavior doesn't surprise me, and there's no love lost if I no longer install it. I got what I paid for. Others get it to work with no problem; I just didn't have time to work with it.

Hmmm...I built my own PCs in the summer of 2091, and both give me flawless rips as long as I use competent software. I use mainly CD Architect or Soundforge for ripping now, but had the same results with Nero and Roxio. And this was regardless of which drive I used (which was either one of two DVD-ROM drives or my CD-RW, which rips at 25x). Only EAC gave me problems...and CDex worked, but was painfully slow. I've even used Roxio's CD Copier in a pinch, from my DVD-ROM drive to the CD-RW, on the SAME IDE channel. No problems there either. (And even Roxio says it can't be done! )

In contrast, I have an old Gateway Pentium Pro 200 (basically a Pentium II minus the MMX) that I bought with a SCSI CD-R drive installed. Never could get a rip out of that drive without either a click at the end of each track, or mistracking. Only on a brand-new spot perfect CD could I get a clean rip. And I sometimes had trouble with the blazing fast 2x burn speed and had to burn at 1x most of the time. The 8X CD-ROM drive ripped better, but I still had to eyeball the tracks in an editor for random glitches. My newer computers I can trust to give me flawless rips each and every time. And nowadays I never turn off anything during a burn--whatever I'm running (e-mail, browser, Photoshop, FTP, Trillian, Proxomitron, AVG, etc.) is left on, I continue working, and I get flawless burns each time. I've only "coastered" some CDs on one of my computers due to a failure of the IDE controller on the motherboard. I also use WinXP (and recommend, at bare minimum, any NT-class operating system like Win2K)...I don't trust Win98 (or WinME) for any critical work these days. Too unstable.

Bottom line: I burned a few dozen CDs on my old computer, most of them being glitch-free. I've burned hundreds on my newer computers, with only a few glitched CDs. (Aside from a small pile of coasters when that #$&@ $# IDE interface crapped out on me.)

I'm on this side of the fence as well. My computing / programming / logical sense makes me lean this way.

I agree that the term "jitter" is misused in the CD ripping programs out there. I always thought of jitter as an A/D or D/A clocking issue, not an indicator of read errors from a CD-ROM drive (which is actually how they seem to be misusing the term "jitter" in some of these software apps). IMHO, installing a large application like Photoshop or MS Office from CD is the same as ripping an audio CD--you're pulling 1's and 0's from the CD, nothing more. Thing is, when you're installing software from a CD, one misplaced 1 or 0 in an application can create a corrupted file that would cause a program to crash, if it even ran at all.

Now my head hurts...

 

"Please explain how n+1 effects can effect burning when the data file containing the music does not have timing data, only musical content. "

Two ways to introduce jitter even with perfect duplication of ones and zeros:

1. The masterclock is not precisely aligned with the final music data file which causes jitter on playback.

Michael, to be more specific there are timing issues that occur when the clock signal and music content meet which is essential for playback. Even the smallest variations can lead to smearing effects on the sound.

2. Pressing-induced jitter in the disc manufacturing process introduces time distortion.

"I always thought of jitter as an A/D or D/A clocking issue"

This is correct but jitter also occurs elsewhere in the chain as well. See above comments.

"IMHO, installing a large application like Photoshop or MS Office from CD is the same as ripping an audio CD--you're pulling 1's and 0's from the CD, nothing more."

A common belief but entirely wrong. In audio it is a different game altogether which is why some many EEs and software engineers get it wrong. Not only do the ones and zeros need to be correct but they need to arrive at the precise moment in time as originally sampled. Even the slightest 20 picosecond difference in timing causes audible distortion. This is why many recording engineers buy external master clock that exhibit lower jitter (as my team does) and also are meticulous about who makes their discs like Michael Bishop who discovered pressing issues with his superb Telarc releases.

 

Not when ripping. When ripping, the time that each arrives is irrelevant. As long as each bit arrives correctly, and is properly recognized as the 'next bit' and gets written down in the file in the correct order.

Lee, please read again what he said when you replied with this statement:

He said ripping. He didn't say anything about feeding a DAC or real-time copying. He said ripping.

N+1 jitter effects are not relevant to a two-step rip-burn process. Unless you aren't ripping correctly.

My position remains that with the right drive, with your PC set up correctly, jitter in your CD-Rs can only come from the burning side. The rip process and the file on your harddrive become the ultimate, 100% perfect reclock buffer.

 

While I agree with Michael (St. Clair, that is), in that any problems with jitter in a computer-based burn come from the burn, not the rip, the "audio as data" theory isn't totally true. There are two main reasons for this:

- audio on CD is stored as one continuous stream, rather than in files/blocks. Older drives had a hard time accessing the correct spot in the stream on subsequent reads, which caused errors. This has been fixed by better drives and "overlap" functions in software - the software will read slightly more than it needs to to slightly overlap the "samples" it reads. To illustrate, let's say the drive reads 6 characters at a time. Without overlap you might get:

"audio " "on CD " "D is s" "red as"

Certain characters are missed, and some are read more than once. Overlap fixes this:

"audio " " on CD" "D is s" "stored"

- data on CD has quite a bit more error correction than audio. That's why you can't fit 80 minutes of WAV files on a data CD.

 

lukpak,

Right on all counts. I was only looking at it from a jitter perspective. From a data format and access standpoint, there are substantial differences between redbook audio and iso/udf/whatever data file formats.

 

Correct. BTW, I made your first point earlier. I guess no one paid attention...

 

It's not that it's a PC, it's more to do with software issues.

 

"Not when ripping. When ripping, the time that each arrives is irrelevant. As long as each bit arrives correctly, and is properly recognized as the 'next bit' and gets written down in the file in the correct order."

Michael,

You are misquoting me - I agree on the ripping part of the action. Read my post of 11:54AM carefully I am referring to burning of discs as the thread originally stated where there are comments to both ripping and burning.

You said: "Please explain how n+1 effects can effect burning..."

...and I did as clear as possible.

Ripping music off a disc is not an issue at the read stage since the time code has been locked in. It is in keeping that time "alignment" perfect thru to the new copies where things get goofy (mis-timed).

And as I said before, error correction is largely a non-issue as algorithms and checksums fix it effectively.

 

"The rip process and the file on your harddrive become the ultimate, 100% perfect reclock buffer."

Sorry but one more correction on the drive aspects. Bob Katz has discovered differences in hard drive write speeds from 4X improving to 1X in sound quality. He has found sound quality differences with the SCSI drives on the Alessis as well. I trust his judgment having worked on several albums with him but I find the Alessis to have superb sonics overall and it is standard issue for our team.

This may be a matter of opinion but usually Bob finds things before others and his ears are impeccable in terms of hearing.

 

To be exact, computer data uses three forms of error correction while CD audio data uses two.

 

1X and 4X hard drive speeds? Riiiight. Please explain just what '1x' and '4x' hard drive speeds even are.

Lee, I honestly have no idea what you are talking about. Easily half of what you say seems like a complete non-sequitur...sort of beating the wrong dead horse.

Various posters say they agree with me. Not one poster claims to agree with you or even understand what you are talking about.

If anyone can translate Lee-speak into english for me, please come forward...I really need the help! Not a joke!

 

That's true, but such effects are *not* additive (ie, n+1). Each time you do a rip, the timing information on the source is thrown out the window. Thus, you could be on a 100th generation copy, but the amount of jitter on the final disc would only be as much as is introduced in *one* burn cycle.

 

Thanks for the breath of fresh air Luke. You said...

"That's true, but such effects are *not* additive (ie, n+1)."

For the most part this is true, but I am not claiming in all cases the 99th generation copy will be substantially sonically different. I am saying that there exists the possible introduction of additional jitter artifacts if there are extra devices in the chain or if pressing the disc itself introduces jitter. This is the problem that Telarc ran into when it duplicated some SACDs and CDs a year ago. Sony Disc Manufacturing has been working to improve the glass master process as a result.

"1X and 4X hard drive speeds? Riiiight. Please explain just what '1x' and '4x' hard drive speeds even are."

I will skip the sarcasm and respond. These have to do with the write speed on professional CDR recorders (say an Alessis MasterLink). Bob has found that slower speed are much less susceptible to sonic degradation and I believe he attributes this to lower jitter.

 

Are you saying that Telarc uses disc rips as sources for creating glass masters? I'll wager they use master tapes. This is not relevant to the rip/burn discussion.

You didn't say 1x and 4x recorder, you said 1x and 4x 'hard drive'.

Yeah, we've pretty much all said that jitter can occur in the burn process. Certainly none of us have claimed that all recording speeds on a given device will introduce identical amounts of jitter. Of course burned discs can still have pit jitter!

You still haven't introduced any kind of reference to a test-proven incidence of an n+1 jitter effect from rip -> burn CDR copying with proper hardware and configuration. That's what I've been talking about all along.

It is indeed wonderful that a typical $40 consumer PC drive can (when properly configured) extract jitter-free audio, leaving the burner as the only area of concern on that matter.

 

From the Jon Risch article:

He gets it.

 

Bob Katz's conclusions on burn speeds rely on the age of his burner and the media he used. I don't know how much he updates his website, but this sounds like outdated info to me. I do have to disagree with those that maintain that burning 1x is better, in light of changing technology.

 

"It is indeed wonderful that a typical $40 consumer PC drive can (when properly configured) extract jitter-free audio, leaving the burner as the only area of concern on that matter."

Michael,

The problem is that burning almost always introduces sonic degradation of music via jitter. So it is a major concern unless you have low jitter devices.

"So in this regard, computer systems ARE completely different than digital audio systems."

I said that earlier as well. Read my posts. The time code element is missing on computer data and that is why software types and many engineers just don't understand jitter - they reduce it down to a "bits is bits" argument.

"You still haven't introduced any kind of reference to a test-proven incidence of an n+1 jitter effect from rip -> burn CDR copying with proper hardware and configuration. That's what I've been talking about all along."

I am not sure I am going to contribute further as I addressed many times this before. Jitter can be present in the burn process of copies ripped from a CD. Most CDRs, outside of the studio, add significant time distortion. That has been my main point all along.

"You didn't say 1x and 4x recorder, you said 1x and 4x 'hard drive'."

It seemed obvious to me as this is how two different engineering teams I have worked with have termed it.

"Bob Katz's conclusions on burn speeds rely on the age of his burner and the media he used. I don't know how much he updates his website, but this sounds like outdated info to me. I do have to disagree with those that maintain that burning 1x is better, in light of changing technology."

Grant, I can shed some light on this as I had a recent conversation with Bob. We used to work together. He still believes this occurs on even new equipment. We were discussing the convenience of the Alessis and he brought the subject up. He was concerned the 4X may play negative role in our recordings. I have not heard it impact our sonics yet but we have not been able to compare to a slower writer. I defer to his judgment in the interim.