CD: Jitter, Errors & Magic Test gear used

Sidebar 1: Test gear used

This project was made possible by the loan of two pieces of sophisticated test equipment. The first is a CD quality analyzer called the Design Science CD Analyzer, developed by Gordon Rudd in conjunction with Disctronics. It consists of a board that fits in the expansion slot of an IBM-compatible PC, software, and a specially modified Philips CD player. The board takes the HF signal from the player and performs all decoding functions, just like a CD player. Software then analyzes the error types and presents the data in graphic form on the computer display. In addition to measuring error types and rates, the system also measures HF level, tracking level, and asymmetry.

The CD Analyzer distinguishes between each of the CD's error syndromes. These syndromes are called E11, E12, E13, E21, E22, and E23. The designations refer to the severity of errors, in increasing order of severity. E11, E12, and E13 indicate either one, two, or three errant or missing symbols in a block as detected by the first stage of error correction (the C1 decoder). E21, E22, and E23 indicate one, two, or three errant symbols in a block at the second stage of error correction (the C2 decoder). An E23 error is the first stage of error concealment, causing the player to interpolate the missing or errant data.

Test data can be stored on a floppy for later analysis. The system also features a statistical analysis program that compares performance trends over a large number of discs, such as during a CD production run.

The Design Science CD Analyzer is most often used in the quality control stage of CD manufacturing. It is also ideal for record companies who want to verify the quality of CDs made for them by the pressing plant. The Design Science CD Analyzer is available from Design Science, 5245 Sale Avenue, Woodland Hills, CA 91364. Tel: (818) 348-3392.

Jitter measurements were made with a Kenwood DB-3545 Jitter Analyzer, a device used mostly by CD transport manufacturers during quality control. It measures jitter in the HF signal by looking at the time between the falling and rising edges (or vice versa) of I3, the shortest pit or land length on the disc. Ideally, I3 should have a period of 694 nanoseconds (billionths of a second) that remains constant. However, no CD transport is jitter-free, causing the period of I3 to be slightly shorter and longer than 694 nanoseconds. (The eight other frequencies in the HF signal will also be changed.)

The DB-3545 displays the jitter graphically, allowing easy interpretation of jitter distribution and character. Typically, a Gaussian distribution (bell-shaped curve) is seen, as the period of I3 varies around the ideal value. The maximum period of I3 is also displayed continuously in real time with 1-nanosecond resolution. The analyzer also has user-definable jitter limits with a GO/NO-GO display. The DB-3545 measures jitter only in the HF signal from a CD transport, and not in the SP/DIF digital output of a CD player or DAT machine. The DB-3545 sells for $3950 [in 1990—Ed.]. Kenwood makes a complete line of test equipment, including CD and DAT measurement instruments. They can be reached at Kenwood USA Corporation, Communications and Test Equipment Group, 2201 E. Dominguez Street, P.O. Box 22745, Long Beach, CA 90801-5745. Tel: (213) 639-4200.—Robert Harley

NotAvailable's picture

The signal that the photodetector gets is a binary signal: ones and zeroes. So either the photodetector gets the message or it doesn't. There's no 'signal to noise ratio' (well there is, but either it's right or it isn't), the signal is discrete: either it gets a one or it gets a zero. If there isn't, the cd-drive will try to repair the signal with error-correction. It's like saying that you replaced your light switch button, resulting in a lamp that shines brighter...

memanuele's picture

I was just at my friend, T's house tonight and was ready to prove that this was "Complete and utter bull S*&T", so I had him play 2 identical cds simultaneously, One with the "green paint" and one without.  I had my eyes closed so I could not know which one he had selected (both were played on identical cd decks so that would not be a factor).  He would do an "A-B" switch.  I did not know which one he started on, and expected there would be no difference at all.   I was flabbergasted when he threw the switch, and as clear as day was light, I heard was what I can only describe as a "more clear and open sound". Now mind you, I had no idea which one was the clearer sound (the paint or no paint), but there WAS a definite difference.  He even reversed which disk was in each cd player (both very high end players)(as a matter of fact the entire system was VERY HIGH END including Infinity Reference Standard 5 Speakers) and I consistently chose the "painted" cd as the one with the more "open and clear" sound.

Now another friend of mine Roy (not his real name (trying to keep him somewhat anonymous)) (who has a VERY critical ear (he used to be an audio engineer at a Major US record label)) was there and he said he heard no difference.  Now to be fair, he was in a different part of the room than I was, so that may have been a factor.

It was VERY hard for me to describe the difference (all I was able to come up with was more "open and clear") until I read this article (I was looking for some "ammo" to use since "Roy" and I "discussed" this all the way back from T's house till I dropped him off at his house (about 45 minutes away)).  The author of this article's description of "Soundstage depth increased, mids and highs were smoother with less grain, and the presentation became more musically involving" describes EXACTALLY what I heard!!! Now my "ear" may be more "critical" in this nuance of audio only because of my orchestral and choral conducting background, and I know there are times when "Roy" would hear things in other audio tests unrelated to this one that I did not (I want to make it clear that I'm NOT trying to say that he was lying when he said there was no difference) so maybe our ears are more "critical" in different nuances.

Anyway, the difference was SO CLEAR to me that "Roy" suggested he come up with a "double blind" test because I am THAT convinced that there IS a difference.  Now, I would like to see if someone has ever had access to an original digital master tape, and compare that to both the plain CD and the "painted one" and see if there is a difference between the plain cd and the master tape or the "painted" cd and the tape.

My friend "Roy" was also convinced that an audio cd was just a bunch of "ones and zeros” and that the same "ones and zeros" were on both the plain and "painted" cds.  I will perform a test and "rip" both plain and "painted" cds to my computer.  If I get a different crc on the two files, then there is proof that "Something" is happening with the "painted" cd. I will report back here what my findings were back on this site.


Mark (my real name)

velocitymj's picture

I always can tell a CD that I've had for over 20 years, because they all are marked with green paint from the CD Stoplight pen.
I was an adherent.
But I never really heard a major difference.
Finally I realized that I was imagining something minute as to be non-existent, because I wanted to believe.
Today, I went to play Handel's Messiah (Harmonia Mundi) and I was reminded of those days and since I haven't read anything on it in quite sometime, I was wondering what was being said about this "Tweak" now.