Second Layer
I numbered the Layers from an encryption kind of view. First Layer 1 then Layer 2. If you decipher a text you would do it in the reverse direction. First Layer 2 then Layer 1.
The binary XOR (modulo 2) operation is very simple. It is the equivalent to the mod 26 addition in classical ciphers, based on the 26 letters alphabet. It is used in many digital ciphers. Eg. DES, AES, digital ONE TIME PAD etc.
Binary XOR (bit by bit):
0+0=0
0+1=1
1+0=1
1+1=0
If we want to apply the XOR operation to the K4 ciphertext, what key we should use? According to Mr. Scheidt: Something that "an operative in place" would also have available. Some kind of information or key inside a secret document, only the legitimate receiver of the message possesses. If Mr. Sanborn had used a very long key (eg. a running key) for this second stage, and if he would not have left some information about it, it is impossible to solve the message. So what do we have? The Alphabet Panel on KRYPTOS!
Kryptos SoftwareOn the left picture you can see the result of my KRYPTOS analysis software, having done a complete run, XOR´íng K4 with the entire alphabet panel of KRYPTOS. Letter by letter. The results were filtered, respectively sorted using the Index of Coincedence (IoC). A IoC of 0.06077 is extremely high. A text in English language has typically a coincidence index of 0.0667. But why have I dropped OBKR? Because it don´t matter. It seems that OBKR has nothing to do with the ciphertext. Let me explain this.
The result without OBKR
|
C: UOXOGHULBSOLIFBBWFLRVQQPRNGKSSOTWTQSJQSSEKZZWATJKLUDIAWINFBNYPVTTMZFPKWGDKZXTJCDIGKUHUAUEKCAR |
|
K: FGHIJLMNQUVWXZKRYPTOSABCDEFGHINGHIJLMNQUVWXZKRYPTOSABCDEFGHIJLOHIJLMNQUVWXZKRYPTOSABCDEFGHIJL |
|
P: SHPFMDXBSFYLQFIPNVXREPSSVKALSZASWTQSGQBFSKBZWSMZKCFEKBSLHAJGSPYTTGVKPZBQSSZSFSSPFTJWKQDSBCJKR |
The result with OBKR
|
C: OBKRUOXOGHULBSOLIFBBWFLRVQQPRNGKSSOTWTQSJQSSEKZZWATJKLUDIAWINFBNYPVTTMZFPKWGDKZXTJCDIGKUHUAUEKCAR |
|
K: FGHMFGHIJLMNQUVWXZKRYPTOSABCDEFGHINGHIJLMNQUVWXZKRYPTOSABCDEFGHIJLOHIJLMNQUVWXZKRYPTOSABCDEFGHIJL |
|
P: IECRSHPFMDXBSFYLQFIPNVXREPSSVKALSZASWTQSGQBFSKBZWSMZKCFEKBSLHAJGSPYTTGVKPZBQSSZSFSSPFTJWKQDSBCJKR |
Note: If the XOR operation produced a "00000" - Null letter, the cipher letter is inserted in the result instead. Something similar was done at the 5 bit ONE TIME PAD teletype XOR operation during the cold war. They inserted the key letter at this position. LINK: How the Rockex discriminator worked
K4 High IOCApart from the fact that the index is very high, probably the position in which this happens plays an interesting role. Have a look at the left picture. No matter if OBKR appears in the ciphertext or not, the position with the high IoC is always the same. It starts with UOXO ... at the alphabet panel ID row "M". DY A HR (01101) is 5 Bit ASCII "M"! There is an interesting statement from Jim Sanborn: “Miniature was missing the extra L, and Sanborn asked about an M when looking at it”. (From Elonka´s and John Wilsons Raw Dinner Notes 2005)
This leads to a conclusion, I never heard somewhere else. OBKR is not part of the message but part of the message key in front of the message. The Germans did this with Enigma messages and the Allies too. The creator of the message choosed a message key. This key then was part of the encryption procedure, which included the machine key (inner settings) from a secret cipher document. Ed Scheidt said: “How do you deliver a message in such a way that it conveys not only the encrypted data, but also the key?” (Note from Elonka Dunin's 2003 roadtrip)
It makes perfect sense that OBKR is not part of the message, because it would be very unlikely that OBKR is part of the KRYPTOS pattern at the end of K4, if OBKR was part of a two stage highly secure encryption. OBKR is like "debris" in front of the message. From K3: "SLOWLY DESPARATLY SLOWLY THE REMAINS OF PASSAGE DEBRIS THAT ENCUMBERED THE LOWER PART OF THE DOORWAY WAS REMOVED...".
The Germans also encoded their message keys in front of the ciphertext. Mr. Sanborn had a message key and he encoded it independently from the rest of K4 with OBKR. So he could choose OBKR "randomly" or let's say "intentionally". In this way he could produce OBKR deliberately, to obtain this KRYPTOS pattern on the sculpture. The structure of K4 looks now something like this:
|
OBKR UOXO GHULBSOLIFBBWFLRVQQPRNGKSSOTWTQSJQSSEKZZWATJKLUDIAWINFBBERLINCLOCKWGDKZXTJCDIGKUHUAUEKCAR |
There is a statement from Jim Sanborn about the misspellings or errors in the other messages: He said that the misspellings were deliberate, but it was less important *what* they were, but, "it's more the orientation of those letters that's useful there." Later on in the evening, he repeated the point, saying that it was the "positioning" that was important. (Note from Elonka Dunin's 2003 roadtrip) UNDERGRUUND - UNDERROUND ? The letters U and O separating the key(s) from the message here.
The key from the alphabet panel of Kryptos, used in the XOR operation above, contains the extra "L" from the panel. Without the extra "L", we got an Index of Coincidence of 0,04558205. This is also much better than the extreme low 0,03608247 from the original K4 message. Remember the second layer is, according to Ed Scheidt, a "masking technique" to mask the statistics of a maybe poor first layer technique.
|
C: UOXOGHULBSOLIFBBWFLRVQQPRNGKSSOTWTQSJQSSEKZZWATJKLUDIAWINFBNYPVTTMZFPKWGDKZXTJCDIGKUHUAUEKCAR |
|
K: FGHIJLMNQUVWXZKRYPTOSABCDEFGHINGHIJLMNQUVWXZKRYPTOSABCDEFGHIJOHIJLMNQUVWXZKRYPTOSABCDEFGHIJLP |
|
P: SHPFMDXBSFYLQFIPNVXREPSSVKALSZASWTQSGQBFSKBZWSMZKCFEKBSLHAJGSPVTTAWHAKAPDQQJMZWKZFIVLPGRMBIMB |
Letter Frequency with LThe next 2 pictures show the letter frequency of the XOR result. With and without the extra "L". The message now misses the letters "U" and "O". (undergruund / underground ?) The frequency analysis was done with the tool from the rumkin.com website.
Letter Frequency without LIn the Layer 1 section we will see, that it is maybe not necesary to use all the letters from position "M" on the alphabet panel. Maybe we can identify some other rows which may be sense. But the beginning of the "M" row is still useful.
Is that what the mask technique should hide?
-->Next Page<--