Variant Beaufort Cipher

Encryption is a variant of the Beaufort cipher, itself being a variant of Vigenere, so it uses a key (and an alphabet).

Example: Encrypt the plaintext DCODE with KEY as keyword and the latin alphabet ABCDEFGHIJKLMNOPQRSTUVWXYZ

Each letter has the value of its rank in the latin alphabet from 0=A,1=B,...,25=Z

Example: DCODE is first coded to 3,2,14,3,4 and KEY to 10,4,24

Encryption consists in subtracting the key to the plain text. Calculation is made letter after letter (ie. their rank/value in the alphabet).
In case of negative result, add 26 to the result (with 26 the length of the alphabet).
The key is repeated (if needed) until it fits the plain text's length : KEYKEYKEYK...

Example: Subtract the first letter of the key K (=10) to the first letter of the plain message D (=3) : 3-10=-7. As the result is negative, add 26. -7+26=19. Save this result and go on with next letters : the 2nd letter of the plain message C and the 2nd letter ot the key E : 2-4=-2+26=24. Keep going with the third letters O and Y : 14-24=-10+26=16. At the 4th step, arrived at the end of the key, repeat the key (or go to the beginning, its the same), subtract the first letter of the key K (as K is the 4th letter of the repeated key) to the 4th letter of the plaintext D : 3-10=-7+26=19, and to finish E and E so 4-4=0.

Each result is a number from 0 to 25 having a letter of the same rank in the alphabet that gives the cipher text.

Example: 19,24,16,19,0 becomes with 19:T, 24:Y, 16:Q, 19:T, 0:A, the cipher message TYQTA.

Variant Beaufort cipher adds the message to the key, this is equivalent to encoding with Vigenere.

A message encrypted with Variant Beaufort has an index of coincidence oscilliating 0.04 and 0.05, often smaller than the one of the language of the plain text.

Techniques to decode Beaufort are similar to the ones for Vigenere. dCode can analyze the text in order to find the probable key lengths (Kasiski, etc.) and uses frequency analysis techniques to find the key.