Time-based parameters

• JC - Jitter Combined

Combined jitter relative to PLL clock. The PLL is a pulse with a period length equal to T or 38.2 ns, and is generated from a digitized data stream. The digitized data stream will be compared to this PLL clock. All present length effects in the data stream are used to form a histogram (JCH) from which the jitter value is being calculated. The standard deviations or jitter of the measured values are normalized to the PLL clock. The jitter generated by the variance from a DVD to the reference plane is compensated with the equivalent tilting from the pick-up optical axis.

JC is the standard deviation of the falling and rising edge of the digitized and equalized HF signals in respect to the PLL clock.

For pre-recorded media, a high jitter value means that pits and lands are shorter or longer than their specified length. If these variations are above a specific tolerance, read-out failures will be generated because of a wrong decision on the nature of a pit or land.

Jitter is also a rather complicated matter in recordable media. As with other signals, dye properties play a crucial role in the value of jitter of the written marks. In addition, the writing strategy and its implementation in the writer used can influence the Jitter as well. Moreover, phenomena related to the sequence of different recorded spots and lands (spaces) called inter-symbol interference, also influence the length of recorded marks.

The primary feature responsible for jitter in recordable media is the dye itself and its decomposition properties. The actual media jitter is a result of tedious optimization of dye application, and writing strategy formulation suitable for each speed.

Digital Error Parameters

In order to make an error correction possible, redundant data has to be added to the user data. With DVD, the user data are organized to form a matrix. The width of the matrix is 172 columns and the length of each data sector is 12 lines (172 x 12). All together 16 data sectors are dedicated to this matrix, which results in a total of (12 x 16) = 192 lines.

With the assistance of a special algorithm, two sections of parity data are computed out of the original data and added to that original data. First of all, to each of the 172 columns, 16 bytes of PO parity data are generated, through which the matrix is enlarged from 192 to 208 lines. Again, of each of the 208 lines, 10 bytes PI parity data are generated. The matrix is now 182 columns wide.

This enables a player to correct at least five defective bytes in each line and at least eight defective bytes of each column. With the assistance of various alternating calculations over the lines and the columns it is possible to correct more defective bytes.

For the evaluation of the following error flags a summary of eight matrix or ECC blocks is used.

• PI Sum 8 - Parity Inner Code Error

The parity inner code error is a continuously measured value, reported for every megabyte. This is the number of parity Rows, summed over 8 consecutive ECC Blocks, containing any bad bytes.

Such errors can be caused by all optical defects that interfere with the normal data read-out procedure (scratches, dots, black spots, bubbles, etc). Local defects can cause the data contents to deteriorate Others causes can be jitter problems, stamper defects or a bad HF signal.

The other causes for high error rate can be the recording layer itself. For example, the dye layer can contain particles, and causes coating defects which, if the medium is not rejected during manufacturing process, can show a high error rate. Similarly, too thin layer of metal mirror can result in a high value of digital errors.

• PIF - Parity Inner Code Fail

The PIF signal or parity inner code first level fail is a continuously measured value, reported for every megabyte. PIF is the number of rows within an ECC Block with too many bad bytes to allow error correction. All bytes in the row are marked bad.

Such errors can be caused by all optical defects that interfere with the normal data read-out procedure (scratches, dots, black spots, bubbles, etc). Local defects can cause the data contents to deteriorate Others causes can be jitter problems, stamper defects or a bad HF signal.

The signal is not specified in the DVD Book.

• POF - Parity Outer Code Fail

Parity outer failure is a continuously measured value, reported for every megabyte. This occurs when the decoder is unable to correct all the data within an ECC Block after attempting one pass of PI Row correction and PO column correction.

The wording "first pass" is important because the decoder may be designed to make several passes of error correction. There is the possibility that bad bytes may be corrected in one of the subsequent passes. DVD and DVDR Tester decoders use one pass of error correction only.

Such errors may not be correctable and should be avoided on a DVD, since the data integrity is damaged.

Such errors can be caused by all optical defects that interfere with the normal data read-out procedure (scratches, dots, black spots, bubbles, etc). Local defects can cause the data contents to deteriorate Others causes can be jitter problems, stamper defects or a bad HF signal.

References:

DVD Specification for Read-Only-Disc Part 1 Physical Specification version 1.0

Standard ECMA 267 - 120mm DVD Read-Only Disc 2nd ed.

Datarius Group