ASCII85 Encoding

Encryption uses the binary code of the text (which depend on the encoding used: ASCII, Unicode, etc.)

Example: Consider the message DCODE encoded 01000100 01000011 01001111 01000100 01000101 (ASCII 8-bit encoding)

The binary string is then split every 32 bits (4 bytes)

Example: 01000100010000110100111101000100 (base2) = 1145261892 (base10) and 01000101000000000000000000000000 (base2) = 1157627904 (base10) (filled with 3 null characters = 24 times 0 to get 32 bits)

Numbers obtained are then converted in base 85.

Example: 1145261892 (base10) = [21,79,73,64,27] (base85) and 1157627904 (base10) = [22,15,0,26,69] (base85)

Each value in base 85 is replaced by an ASCII character of code (value+33)

Example: 21 => ASCII(54) = 6, 79 => ASCII(112) = p, 76 => ASCII(106) = j etc.

Example: The encrypted message is 6pja<70

There could be some final null characters ASCII(0), that can be deleted (it corresponds to the previously added null characters)

The decryption start by splitting the text into groups of 5 characters (if the tuple is not 5-character long, it is filled with u)

Example: Consider the message 6pja<70, split into 6pja< and 70 (that is treated as 70uuu)

For each group of 5 characters, get the ASCII code and subtract 33.

Example: 6pja< = 54,112,106,97,60 (ASCII codes), -33 => 21,79,73,64,27

Example: 70uuu = 55,48,117,117,117 (ASCII codes), -33 => 22,15,84,84,84

Every tuple of 5 numbers is then considered as a unique number written in base 85, and converted in base 2.

Example: [21,79,73,64,27] (base85) = 01000100010000110100111101000100 (base2)

Example: [22,15,84,84,84] (base85) = 01000101000010010101011000000101 (base2)

The plain text is then this binary code encoded with initial encoding (ASCII, Unicode, etc.), n last characters are ignored when n characters have been added in initial completion step.

Example: 01000100 01000011 01001111 01000100 01000101 (00001001 01010110 00000101) = DCODE (ASCII 8bit encoding).

The message uses only ASCII characters of codes 33 to 117 with sometimes spaces, carriage return, but also characters y and z.

The message begins with <~ and ends with ~> (variant Adobe)

the message begins with xbtoa Begin and ends with xbtoa End (variant BTOA)

Adobe ASCII85 is used in PDF files, it begins with <~ and ends with ~> and authorizes z exception.

The character z is used (even if it is not in the character list) to indicate a group of 0, this trick allow reducing the size of the message. (used in Adobe and BTOA variants)

The character y is used to indicate a group of spaces. (not compatible with Adobe)

ASCII85 code has been created around 1990.