Frequently Asked Questions About Compact Discs

Please contact Media Sciences if your questions are not answered on these pages.

Is it true that jitter is the most important measurement of CD quality?

Clip level returns in the guise of jitter! Jitter is definitely not the most important measurement of CD quality. The most important measurement is always the one you do not perform. Hidden problems are much more likely to adversely affect quality than issues you monitor.

Amplitude parameters such as I3, I11, asymmetry, push-pull, radial noise, crosstalk, and reflectivity are vital. Major timing measurements are length deviation and jitter. Error rate measurements such as BLER, E11, E21, E31, E12, E22, E32, and BURST provide indispensable information. Important mechanical properties include center hole and outer rim diameter, thickness, weight, unbalance, eccentricity, deviation, and deflection. Logical volume and file structures are another important quality requirement. CD quality testers should measure most of the above. Manufacturers use many tools to regularly check every one of these plus others such as birefringence, track pitch, scanning velocity, and radial location of key areas such lead-in, program area, and lead-out.

Diskette duplicators are familiar with window margin, one good indicator of recording quality. This time error represents the sum of two parts, peak shift and bit jitter. Peak shift is the difference between the ideal and actual average time placement of read data bits. Bit jitter is the variation around that average. Values for each are obtained from time interval analysis (TIA) of three MFM lobes.

CD discs have nine TIA lobes for lands and another nine TIA lobes for pits. Eighteen values of length deviation represent the difference between ideal and actual average time placement of read data bits, and are comparable to diskette peak shift. Jitter is the variation around that actual average, and is reported as the one standard deviation value for the TIA lobe. Bit error rates can then be predicted based on length deviation, one sigma random jitter, and player jitter.

All eighteen length deviation and jitter values are critically important to proper control of the manufacturing process. But low length deviation and jitter do not always indicate high quality. Although they can predict low average error rates, they are not very sensitive to local defects such as black spots, birefringence, and scratches. Two levels of cross-interleaved Reed-Solomon error correcting codes easily detect and correct small, random errors. Even errors resulting from a one mm hole in a disc can be detected and corrected. Now that's vaporware! But would you knowingly ship such a disc? This is where proper error rate measurements become important.

Similar examples can be given for other quality issues. Modern CD testers simultaneously measure many quality indicators during one test sequence. Use all of the available information to predict successful interchange. Avoid reliance on just one indicator such as jitter. Your false sense of security may receive an unpleasant jolt when the phone rings!