Re: Cryptomania
 

Good point..... means I somehow screwed up either the letter sequence or the translation - because I did end up with 42 triplets in the end :lol:
Was the rest OK?

I would agree with the 000 though :nod:

Let me think :thinking:

Re: Cryptomania
 

I got these numbers:
201 120 012 020 120 222 210
112 001 200 210 121 121 210
122 001 202 201 000 110 120
011 120 110 011 111 111 201
121 002 101 110 222 210 112
001 202 201 222 110 120 111

...unless I made a mistake, there are a couple of differences with your grid

Re: Cryptomania
 

Yes, I had added an extra U to the string before :biggrin:

[edit] I suppose we can also rotate the other way round ;)
- bring the last 4 numbers forward in stead of the first 4 back

[edit2] Progress: that last option gives

Re: Cryptomania
 

Since I still see the grid here, please recall #978, where I confirmed Hel who found out that it's just the whole string - and not six seperate lines. The grid was necessary only because I had to fit the text in this small gif area.

... we can also rotate the other way round ...
And you really should, if you follow our blonde - after all she's rotating clockwise (aka to the right), isn't she? ;)

... 'hydrogen peroxide' ...
Aha, the first beams of light appear at the end of the tunnel. :lol:

One thing: only now I noticed that I accidentally used an early version of the text subimage (which had a little error in it - a Z instead of a blank) while composing the gif. :dizzy:
It's rather marginal, and you will get the solution regardless, but just in case: this would have been the correct sequence to rotate:

201 120 012 020 120 222 210 112 001 200 210 121 121 210 122 001 202 201 000 110 120 011 120 110 011 111 111 201 121 002 101 110 222 210 112 001 202 201 000 110 120 111

Re: Cryptomania
 

Well, that translates to:



Interesting though, how a rotation per row gave a partial solution, with at least some sort of pattern, in the top half, but only bits and pieces in the bottom half .... :thinking:

Re: Cryptomania
 

... partial solution ...
Well, only the (two) starting letters would possibly suffer from false digits (coming from false source locations) - all the other letters would be (and in fact they are) correct.

:arms: And donate hydrogen peroxide and ammonium hydroxide (for hair bleaching) it is!

You're up, Hel :jossun:

Re: Cryptomania
 

In fact, only the second letter (I redid the row-rotation with your corrected triplet) is affected it seems, the first just got shifted to the row above. Or is that just a coincidence :thinking:



I'll be back - some time :biggrin:

Re: Cryptomania
 

... just a coincidence ...
No. Given the correct rotation flow, a tail would land at the beginning of the next line. If we process lines seperately, the tail just lands at the beginning of the same line (instead of the next). That's why the starting letters are displaced by one line. ;)

Interesting sidenote: If we look at the 2nd letters, we know that - after 3 steps - they've got exactly their predecessor's (the 1st letter) content. Then, the 4th rotation brings a digit which comes from an incorrect source - the line's own tail.
However, this 'incorrectly delivered' digit (either 0, 1 or 2) might be still correct - with a probability of 1/3.
Funnily enough, the statistics proves true in our case: 2 out of 6 ...
y f n a t e h
p d r o g e n
e e r o x i d
o j n d a m m
d w i u m h y
d i o x i d e
... 2nd letters are indeed correct. :lol:

Re: Cryptomania
 

Ah yes, of course; but it was indeed a lucky coincidence that the '2nd letters' inside hydrogen peroxide were correct :biggrin:



Next:

Re: Cryptomania
 

I opened the spoiler....
Attachment 74563

It says "sorry, this person ate the image for dinner" or something like that....

Re: Cryptomania
 

Haha, yes. I made a slight change to the pic, to make it even more simple to solve, and when I uploaded the new version, Photobucket died on me (or had gone out to dinner maybe?)

Here it is again:

Re: Cryptomania
 

Do we need this...
Attachment 74586

... or is a normal phone ok?

Re: Cryptomania
 

Normal phones should work :lol:

Re: Cryptomania
 

Hmmm... already tried normal phones; at least I know which buttons he presses. Couldn't convert the resulting numbers back to readable text...
There seems to be an intermediate encoding I haven't thought of yet :confused3

Re: Cryptomania
 

Each 'statement' only has a single encoding ;)

Re: Cryptomania
 

Haven't investigated much in the meantime. Bet I'm overthinking anyway, as always. Just for the record, here's my list from yesterday (the numbers to be pressed on a phone, to get the ciphertext letters):
555 55 33 5 999 33 999 5 999 55 999 888 2 66 4 9 6 4 7777 9 555 55 44 555 55 999 9999 3 888 666 8 3 77 88 22 555 444 222 666 9 555 5 99 44 2 44 4 88 888 33 7777 777 888 55

... 'statement' ...
Huh? You mean 'word' or 'syllable'?

Re: Cryptomania
 

No, simply each of the two 'pieces of text' shown in the image ;)

Not so much overthinking, as going off into the wrong direction I think, though your basic idea was right :nodder:

Re: Cryptomania
 

... two 'pieces of text' ...
Ah, ok.
Well the top text was easy ("E.T. phone home!" in base64), but that's what I thought of anyway when I saw the pic the first time. So it wasn't much of a help - except the blue colour and underline have an additional meaning (besides emphasizing the words 'phone' and 'home').

The left statement, however, didn't reveal striking patterns so far. Tried grouping (2, 3, 6, lines, columns) in combination with several encodings. Bet I'll scream loudly when we find the solution is totally simple :lol:

Re: Cryptomania
 

There was a reason for the colours. One you picked up on, the other you didn't ;)

Re: Cryptomania
 

... the other you didn't ...
In case you're referring to the 'home' part... already tried including his home, Brodo Asogi (aka the Green Planet) in the Perrinn sector. To no avail. Maybe we need the exact coordinates :idea:

[EDIT]
According to several sources, grid coordinates: J-4
And other sources: (-111, 342)

Re: Cryptomania
 

Now you're overthinking :biggrin:!

Re: Cryptomania
 

I don't get the clue about "home", either...
But going back to the "phone" part, you said Bolaf's basic idea was right. My first idea was not repeating numbers, but making pairs: Push 5 three times, then 5 twice... that is:
5-3 5-2 3-2 5-1.... and so on.

Re: Cryptomania
 

Nope, that's not it either, you're applying the 'number conversion' to the wrong part...

Re: Cryptomania
 

... applying ... to the wrong part...
Ha! I knew, this zigzag frame would have a meaning. Those 21 lines just cannot be coincidental :lol:

Re: Cryptomania
 

:lol: Speak about overthinking...... that's just my effort to draw a 'spark', to show that that part was the message - and the other (base64) text is what he was thinking ;).

Re: Cryptomania
 

... that part was the message ...
I never had any doubt about that part to be the message (= actual ciphertext), with or without decoration :rolleyes:
Or do you mean a message for him? That is, someone called him?

Re: Cryptomania
 

No, he makes the call. But the fact that the spark had 21 sides has no relevance at all ;)

Re: Cryptomania
 

Still wondering how Bruja was applying the 'number conversion' to the wrong part...

Based on the information we're given...

• The algorithm uses a 'single encoding'.
• It probably has to do with a phone numpad (usually 2=ABC, 3=DEF, 4=GHI, 5=JKL, 6=MNO, 7=PQRS, 8=TUV, 9=WXYZ)
• Both, ciphertext and cleartext, consist of letters: all in all a letters->letters (maybe letters->numbers->letters) conversion.

... it's an obvious idea to translate the letters to numpad-values (as Bruja and I did), intending to convert those values back to cleartext (using a common numbers->letters rule such as index/position, ASCII or similar).
Alternatively, one could apply a common encoding (index/position, ASCII, etc.) first and then interpret those numbers as numpad-values, which leads to letters as well.

A little problem I faced here, is, that index numbers and numpad-values can't be directly mapped to each other, because their digit ranges are different - basic numpad-values only contain [2...9], since the buttons '0' and '1' don't possess letters (although this fact might be used to express blanks, though).

In case the algorithm does not use an intermediate number-conversion step, I can - as of math related encodings - only think of base64 and base36 to come directly from letters to letters. Outside math there's of course a bunch of other methods, such as 'drawing letters on the numpad' or similar ways that exploit visual shapes of symbols (aka Bruja-esque).

Also, the grid might contain meta-information, e.g. groups, vertical lines, rotations, coordinates or visual mapping.

So far, I delved into each of those issues. Not too deep, at least you signified it's easy to overthink.
Some observations:
- The message contains 24 (out of 26) distinct letters, so probably not a vanilla substitution cipher (Except you've used a bunch of rare letters, like, P V B G K J Q X or Z. However, I wasn't in the mood for bruteforcing tests either).
- Building pairs/groups makes the stats (for simple substitution) even worse; and you didn't approve that idea anyway. So it's probably not about grouping (which, in regard of digits, held the possibility of building unicode values).

Luckily, while writing the above text, I got some other ideas :) ... must check that out...

Re: Cryptomania
 

- phone numpad and lettering (2=ABC, 3=DEF etc) is correct
- letters to letters decoding of the ciphertext is all you need there
- no rearranging of the ciphertext is needed (9 x6 grid is just for esthetics/space)
- not sure what 'vanilla substitution' is ;)

Re: Cryptomania
 

So... while working with the message we have to translate -somehow- from letters to letters without taking into account any numerical values...
...And then apply some kind of number conversion to another part of the cryptogram.

...I'm about to cry :frown:

Re: Cryptomania
 

@ Bruja:
The number conversion (here: base64) had to be applied to ET's thought.

Re: Cryptomania
 

I wasn't talking about the base64 conversion ;)

I didn't exactly say that.....

No, that's the first step, or you won't be able to 'translate' :)

Re: Cryptomania
 

So, before we attempt to translate the message we have to do something with "home", for instance... or home+telephone, or ET's thought as a whole, right?

Re: Cryptomania
 

Yes, that's it (that's why I used the colour blue in his thoughts ;))

Re: Cryptomania
 

Apply "phone" code to "home"? As in h=44 o=666 ...?

Blue? is blue important? couldn't it be any colour???

Re: Cryptomania
 

Yes :nod:

no :no:

Any colour you want, as long as it wasn't black (as Henry Ford didn't say :biggrin:)

Re: Cryptomania
 

Oh, some more hints appeared in the meantime... doesn't matter, I tried bruteforcing and finally got this: hey guys guess what i am the big star in a movie with drew barrymore.

Re: Cryptomania
 

How do you bruteforce something like that :lol:???

It's the right answer anyway, now the method....

Re: Cryptomania
 

... How do you bruteforce ...
As explained earlier, I checked lots of the usual encodings and also some substitutions, but the statistics alone always said 'sorry, very unlikely'.

After a while, I got the feeling that I must have misunderstood your hints at some point. So I completely restarted my analysis, ignoring your info about 'single encoding', and (just for fun, actually) assumed that the ciphertext was encrypted instead - in other words, a key had been used.
So I performed some letter frequency analysis and - surprise - found a high probability for a key of length 4 or 8.
Then I applied some tests (using XOR and Vigenere) where I checked for common short English words, such as 'the', and suddenly fragments of 'star', 'guys' and 'movie' also appeared rather quickly.
As for the method, in the end it was a vigenere cipher using the key 'eggd' - a strange word; but now, that I read your recent posts, I suppose it's been derived from the word 'home'. I'm not going to do reverse engineering (to come from 'eggd' to 'home') now, I think that Bruja is working on such a step anyway.

[EDIT]
Ok, found the home->eggd method:
h = (on pad button) 4 = (index of) e
o = (on pad button) 6 = (index of) g
m = (on pad button) 6 = (index of) g
e = (on pad button) 3 = (index of) d

Re: Cryptomania
 

There's a difference between the words 'encoding' and 'encrypting'? If so, I never thought about that, I use them interchangeably.

What I used to encode/encrypt is in fact a Gronsfeld cipher. It relies on a numerical key (here: home=4663) and then replaces the letters by, in this case, those 4, 6, 6, and 3 positions 'further on' in the alphabet; in that way I guess it works the same as the Vigenère cipher with keyword 'eggd'.

Couldn't think of a way to put 'Gronsfeld' in the drawing without giving it away immediately, but I figured that once you'd made the connection phone->letters on the phonepad->numbers, it wouldn't be too hard to think of a way to use '4663' with the coded text to decipher it ;). Hats off for doing it the hard way though :nod:

Re: Cryptomania
 

... difference between the words 'encoding' and 'encrypting' ...
:lol: Oh yes... here's a quick recap:

Encoding (e.g. ASCII, Morse, Braille, base2...base64, atbash, substitution/cryptogram, steganography):
- no key
- changes the symbol set (e.g. letter <-> decimal <-> binary <-> RGB <-> visual pattern <-> sound)
- preserves the symbol meaning (e.g. the number 9 retains its meaning; whether noted as "i" or (bin) "1001" or (in Morse) "----.")

Encryption (e.g. Caesar, Vigenere, Playfair, DES/AES, PGP):
- uses a key
- usually preserves the symbol set (e.g. letter<->letter, byte<->byte)
- changes the symbol meaning (e.g. "a" gets transformed into "c", hence this "c" now means "a"; moreover, each "c" can have a different meaning)

Addendum: Outsiders often use those expressions interchangeably (as codes and encryption stick tightly together by nature), though by now one might rightly state we're not innocent outsiders any more... ;)
Likewise, the term 'cipher' is frequently used for 'code' as well, while on the other hand, the word 'ciphertext' doesn't explicitely mean 'encrypted via key'; the term is widely used for all kinds of unreadable or cryptic looking code, e.g. for Poe's Gold-Bug puzzle - which is, as we all know, but a pretty visual encoding (ordinary 1:1 substitution cipher).

... Gronsfeld ...
... once you'd made the connection ...
:nod: That was a fine logical puzzle. In retrospect I can only blame myself for letting me distract by vague assumptions and believing on insufficient evidence. Anyway, I love such piles of stacked challenges, fanciful concepts and (sometimes self-made) red herrings...

[EDIT]
Speak of the devil... look who's there:

----

Attachment 74810