Method and apparatus for reducing DC components in a digital information signal

In a method of recording a digital information signal comprised of a plurality of N-bit data words by means of a recording apparatus which imperfectly transmits DC components of said digital information signal, the steps of:

• selectively generating M-bit data words in which M is an even integer larger than N and each of said N-bit and M-bit data words is comprised of a plurality of logic level "1" bits and logic level "0" bits;
• weighting each of said M-bit data words on the basis of a first value corresponding to a number of boundaries between adjacent "1" and "0" bits in the M-bit data word, a second value corresponding to the number of said "1" bits in the M-bit data word, and a third value corresponding to the numbers, in the M-bit data word, of isolated "1" and "0" bits and of isolated pairs of "1" and "0" bits bounded on both sides by "0" and "1" bits, respectively;
• grouping the weighted M-bit data words in accordance with the run length of data thereof;
• classifying the grouped M-bit data words into a first group consisting of said M-bit data words which have an equal number of said "1" and "0" bits, a second group consisting of said M-bit data words which have a number of said "1" bits greater than the number of said "0" bits, and a third group consisting of said M-bit data words which have a number of said "0" bits greater than the number of said "1" bits;
• assigning each of said N-bit data words to at least a respective one of said M-bit data words in accordance with the generation frequencies of said N-bit data words so that the run lengths of the assigned M-bit data words will vary generally inversely in respect to the generation frequencies of the respective N-bit data words, with some of said N-bit data words being assigned alternately to said M-bit data words of said second and third groups and the remainder of said N-bit data words being assigned to said M-bit data words of said first group;
• converting said N-bit data words to said respective assigned M-bit data words so as to minimize DC components of the digital information signal, as converted, and to concentrate its frequency spectrum;
• and
• recording on a record medium said M-bit data words to which said N-bit data words have been converted.

Apparatus for reducing DC components in a digital information signal to be recorded and comprised of a plurality of N-bit data words by converting said N-bit data words to respective M-bit data words at least some of which are assigned thereto in accordance with the generation frequency of said N-bit data words and the run length of data of said M-bit data words, comprising:

• output means for receiving said M-bit data words;
• memory means for storing said M-bit data words either in the actual or inverted form thereof and which are read therefrom in response to said N-bit data words;
• each of said M-bit data words being comprised of a plurality of logic level "1" bits and logic level "0" bits and being stored in said memory means with the number of said "1" bits being greater than or equal to the number of said "0" bits;
• transmitting means connected with said memory means for transmitting said stored M-bit data words read from said memory means to said output means;
• inverting means connected with said memory means for inverting said stored M-bit data words read from said memory means and supplying said inverted M-bit data words to said output means;
• and
• gate means selectively enabling one of said transmitting means and inverting means to supply said transmitted M-bit data words and said inverted M-bit data words, respectively, to said output means, said transmitting means being enabled whenever an M-bit data word read from said memory means has an equal number of said "1" and "0" bits and said transmitting means and inverting means being alternately enabled when M-bit data words successively read from said memory means have unequal numbers of said "1" and "0" bits.