Eight-to-Fourteen-Modulation

Eight -to- Fourteen Modulation ( EFM) denotes a line code, which for recording data on optical media such as compact discs ( CDs), mini discs and laser discs is used. The EFM method and its extension EFMPlus was developed by Kees Immink and A. Schouhamer is defined in industry standard ECMA -130.

Description

Each byte (data block consisting of 8 bits) is doing a 14 -bit long code word of a table ( defined for CD - Audio in the Red Book, complies with EN 60908, and CD -ROM Yellow Book corresponds to ECMA -130 ) is assigned. Then separating the so-called three - bits are inserted between the 14 -bit code words.

On a CD, the data are stored without depressions (lands ) in a track in the form of depressions (pits ) and intermediate track sections. A short section of track about 1/3 micron length ( a bit cell ) corresponds to one bit. A bit cell with a transition from pit to land or land to pit this case represents a 1, the bit cells in between represent (without transition) zeros (NRZ -M encoding). This can be a CD reliably produced and played (so that the pits and lands are long enough to be reliably detected by the laser ), between the ones least 2 and may be located more than 10 zeros ( so-called d / k- condition) must. Through a code that satisfies the d / k- condition needs a bit cell only to have 1 /3 of the length that they would have if two ones (two bit cells, which have a transition and thus the beginning and end, ie the length of the pit or the country determine ) should immediately follow each other. Although more than twice as many bits need to be stored, due to the fit shortened to 1/3 length of a bit cell in the results of 50% more information on the CD.

At 14 bits used, there are 214 = 16384 among the 267 possible combinations which meet the d / k-constraint.

To ensure that (ie at least three and not more than eleven Bitzellenlängen between two transitions ) is also maintained in the transition region between two 14- bit codewords of minimum distance of two and the maximum distance of 10 zeros are between the codewords in addition three so-called Trennbits (000, 001, 010 or 100) inserted. As in the cases where 10 zeros at the end of a 14 -bit codes, and 10 zeros at the beginning of the next 14 -bit code Trennbitkombinationen by any of the three above, the D / K condition can be ensured, to the above-mentioned 267 existing code words those are eliminated with 9 and 10 zeros at the beginning and those with 9 and 10 zeros at the end. These are 11 combinations, so just every 28 = 256 different values ​​of the 8 bits of the bytes to be stored can be represented by a combination of each.

Such a coded signal would - have times sections with more lands and times sections with more pits - randomly. This change would also create undesirable low-frequency signal components in the high-frequency signal, which would disturb the tracking mechanism of the CD player. The occurrence of these low-frequency signal components is suppressed by corresponding Trennbitauswahl in coding, but this is possible only if the d / k-constraint allows alternative Trennbitauswahl, but this is sufficiently covered by the four possible Trennbit words.

In many cases, in fact, it is possible to choose one or the other Trennbitkombination. For example, I can choose three zeros at the end of the codeword, and three or four zeros at the beginning of the next codeword each of the 4 existing Trennbitkombinationen. Is chosen which leads to a suppression of the low frequency signal components. In this case, you simply calculated continuously from the first code word how many bit cells are on the lands, and pulls the bit cells that are in the pits, of which from. The result ( the so-called running digital sum value " DSW " ) is after the first code word somewhere between 8 and -8. Then you calculate using the acceptable in a given situation Trennbitkombinationen (d / k-constraint ) as the DSW would develop until the end of the next code word. Are selected from among the alternatives that Trennbitkombination wherein the DSW is as close as possible to zero. So goes one and always chooses the Trennbitkombination that brings close to zero the DSW. Thus there are never times more lands and times more pits, causing low frequency signals, but always equally many lands and pits, and thus a low-frequency- free signal (also known as DC- minimized ).

The EFM modulation is described in the Red Book and the Yellow Book in the section " Modulation System '. In EN 60908 (corresponds to the content of the Red Book) is described in Chapter 13, " EFM " and 15 " EFM modulator ". In the ECMA -130 Standard ( the same content as the Yellow Book, on this point, however, the Red Book) it is described in section 19.1 "8- to- 14 Encoding" and 19.3 " mergin bits " and in Appendix E " mergin bits ". The EFM was, for example, by the German patent DE 3119529 patented ( by time off).

EFM

Eight -to- Fourteen Modulation Plus ( EFM ), the modulation used for DVDs of the data stream. Are recoded 16 -bit code words and then recorded 8- bit bytes. The word Fourteen in the name is only for historical reasons at the EFM, because 14 -bit code words do not matter here.

The polarity of the signal must be here every 2 to 10 cycles (d / k-constraint, see EFM) change. Different coding tables, depending on the previous codeword ensure that this happens for all combinations of neighboring codewords, and that the signal spectrum contains few low-frequency components that would interfere with the tracking. The process is more complex than EFM, but only 16 instead of 17 clock cycles for 8 data bits are required, ie it can fit about 7 % more original data onto the disc

In 216 = 65536 possible combinations, there are 566 16- bit code words which meet the d / k-constraint. From this one then eliminated those. 10 zeros at the beginning and those who have 10 zeros at the end Then one can generate four coding tables, each of which includes 256 different code words, which may represent the 256 possible bytes. Here, in the first encoding (status 1) the 256 code words with only 2-9 leading zeros before, in the second and third coding (status 2 and 3) 256 code words with only 0-5 zeros at the beginning, and fourth in the coding ( Status 4 ) 256 code words with only 0 or 1 zero at the beginning. In many cases they overlap, that is, the same codeword may, for example, are both in the state -1 as well as in the status -2 encoding, but then represents the same byte. The same applies to the code words in the state 3 and state 4- polarizing. The state -2 and state -3- coding tables are manufactured so that they do not overlap. All codewords in the status -2 encoding at bit positions 3 and 15 have each a zero, all codewords in the status -3 encoding at bit positions 3 and 15 have at least a one. Using this feature you can find in the beginning, whether they come from the status -2- or -3 status table in the code words with 0-5 zeros.

Depending on the number of zeros that has a code word at the end of the next code word in the corresponding status encoding is selected. Ends the last code word, for example, on 6-9 zeros must be selected the next codeword from the Status-4 - encoding, in which only codewords with 0 or 1 are zero at the beginning, it ends eg to 0 or 1 is zero, the next codeword from the status -1 encoding must be selected, in which only codewords with 2-9 zeros are at the beginning. This ensures that there is always at least 2 and no more than 10 zeros followed by a one (d / k condition). This represents a 50% increase in data density safely (see EFM). However, this ensures not even the suppression of low frequencies, which are undesirable because they interfere with the tracking mechanism (tracking).

The special feature of the code is the code words that end in 2 to 5 zeros. These codewords are in the same state - coding each twice and thereby represent two different (8 -bit ) bytes, so they are not unique. Therefore, a decoder, when he meets in the data stream on one such codeword, not this decode without additional information that indicates which byte is as represented by this ambiguous codeword. This information is given in the following code word: Since these ambiguous code words end 2-5 zeros, each case must be subsequent codeword 0-5 leading zeros have (d / k- condition), either from the status -2 or the status -3- coding originate. Comes this follow-up code word from the status -2 encoding ( ie, it has at bit positions 3 and 15 respectively, a zero ) the ambiguous preceding code word represents a byte, otherwise the other.

Through this double occupancy of codewords are each at least 88 code words in all four encoding tables which are actually included in four so-called Alternativcodiertabellen and thus 88 ( of 256 possible ) bytes represent once in all 4 status ( here are not represented two different bytes by the same codeword but the same byte by two different code words). The code words of the Alternativcodiertabellen are selected so that the running digital sum value (see EFM) adversely affect to the influence, the code word has the same status Hauptcodiertabelle. In the encoding, in cases in which the code word can be alternatively selected from the main or Alternativcodiertabelle selected so that the running digital sum value remains as close as possible to zero.

In the remaining 168 bytes, for which there is only the main table and no alternative table, the digital sum value (and thus the low frequency signal component ) can still be influenced by the fact that it is sometimes permissible, the place of a code word from the status -1 Table code word of the 4 - state table ( or vice versa) to be used. This is allowed only if the number of zeros at the end of the preceding codeword and the number of zeros at the beginning of the next code word not total violate the d / k-constraint ( minimum 2, maximum 10 zeros), which must be examined case by case. In these cases, the code word is used in the coding, which holds the digital sum value as close as possible to zero.

The signal kept as free as possible of low-frequency signal components - Both measures will - as in the EFM modulation.

The EFM modulation is in the default " DVD Specifications for Read-Only Disc, Part 1, Physical Specifications" of the DVD Format Logo Licensing Corporation ( DVD FLLC ) in chapter 3.3 "Modulation Method (8/16 modulation ) " as the default for all DVD discs standardized. In the content same standard ECMA -267 described "Suppress control of the dc component" in Annex G "8- to- 16 Modulation with RLL ( 2,10 ) requirements" and in Chapter 22. In brand-new DVD -R disc, the so-called "Control data zone" in the EFM format available. This is detailed in section 3.4.1.4 of the standard "DVD Specifications for Recordable Disc for General (DVD-R for General ), Part 1, Physical Specifications" of the DVD Format Logo Licensing Corporation ( DVD FLLC ). There, the EFM modulation "Modulation Method (8/16 modulation ) " is specified in the same way as in the DVD standard in chapter 3.3. The EFM modulation is, inter alia, patented by the European patents EP 0 745 254 and EP 0789910.

Example EFM Plus

For example, 00000101 and 00000110 correspond in state 1, however, the word 0,010,000,000,100,100th 00000101 in state 1 requires as next state 2 wherein 00000110 requests as a next state 3. The word 00000111 is encoded in state 2 to 0100000000010010 and in state 3 to 0010000001001000. In decoding, therefore, in a case to 0010000000100100 00000101 and in the other decoded to 00000110.