Guild Wars 2

Hey,
i also think, that ArenaNet put a message into this code and i don’t think, that this is just for decoration!

But i believe, that the code-message isn’t the only thing we get from the screen!
I recorded the sound the activated screen produces, because i thought there could be a message in it, too.

I slowed it down, i changed the pitch level, and i really think there could be a heavily alienated spoken message behind all the beeping and rustling. But my audio skills aren’t good enough to do this job right. Maybe here is someone, who want to follow this path.

I added the original recorded track to my dropbox. It’s just an extract, so it could be possible that you have to record a longer one.
You can find it here:
https://www.dropbox.com/s/xgwd0zw6dctgg15/scarletconsole.wav

TaCktiX.6729:

I’m doing a frequency analysis right now, and it’s matching up fairly clearly with normal written language.

Can you post your results even if they are incomplete ?

hungaryan alphabet have 44 letters. so the 42 is not as much as you think.

Has anyone ever checked what kind of configuration this setup follows? Is vertically aligned, or is it horizontally aligned?

Cormac.3871:

I spent five minutes staring at the code trying to find the word “mordremoth”, which should be identifiable as it has 3 letters that repeat themselves. I found something that could be “mordremath” (2nd line of that two line phrase) but that was the best I could do.

:) I’d assume you misread and the actual message you saw was “murder math” but it was in fact just your subconscious speaking to you. (sorry, couldn’t resist)

TaCktiX.6729:

I think she wouldn’t pop in to say “+1” if it was wild speculation with absolutely no basis.

Oh, I very much assume she would!

Yet, still an absolutely fun topic. Regardless if there is a message to find or not, I hope the devs realize that if they just give us such small pieces of enigma once in a while then there will totally be people willing to spent their time solving the riddle, as far stretched as it may seem.

~MRA

Another avenue you guys could consider is that it is instructions — or — status codes.

It’s too late now (unless someone can get online before the patch), see if they do anything when the marionette spawns.

Advent Leader.1083:

Has anyone ever checked what kind of configuration this setup follows? Is vertically aligned, or is it horizontally aligned?

Well, it scrolls vertically, so if it were real language text I would argue that it pretty much would have to be written horizontally (otherwise there’d be no clue or obvious reason to move columns before text there had scrolled away and been lost). But that doesn’t work as an argument here; if we assume it actually does have meaning for us to find at all, the dev’s could have done just about anything.

Darn it, now I’m interested. That’s my afternoon gone.

I can’t stop looking at this as braille . Although it most likely is not, there is very strong similarity in the two. Where as lines would be connected dots.

*edit added attachment

Bazto.9052:

I can’t stop looking at this as braille . Although it most likely is not, there is very strong similarity in the two. Where as lines would be connected dots.

It does look incredibly promising :O I just took a rather quick glance but it does seem to fit the cypher very well!

Edit: I just noticed the name of the picture, is that Braille?

Second edit: I must be half-blind, obviously it is Braille since you said it is, sorry :P

The Kurt.2518:

Second edit: I must be half-blind, obviously it is Braille since you said it is, sorry :P

Oh, the irony!

I was wondering if the black lines were hiding letters like the attached illusion….. but I haven’t had time to play with it myself. o.o

khani.4786:

I was wondering if the black lines were hiding letters like the attached illusion….. but I haven’t had time to play with it myself. o.o

What black lines are you refering to? The spaces between the symbols or what? :P

Well looks like theres a thing on reddit about it: http://www.reddit.com/r/Guildwars2/comments/1w0782/not_spoiler_scarlets_code_alphabet/ Haven’t read through it yet to see if they’ve cracked it or not.
If someone is better at following those reddit things (got to it somehow from this one and searching that_shaman’s posts(?)), can you post their answer?

Nasedo.5349:

Hey,
i also think, that ArenaNet put a message into this code and i don’t think, that this is just for decoration!

But i believe, that the code-message isn’t the only thing we get from the screen!
I recorded the sound the activated screen produces, because i thought there could be a message in it, too.

I slowed it down, i changed the pitch level, and i really think there could be a heavily alienated spoken message behind all the beeping and rustling. But my audio skills aren’t good enough to do this job right. Maybe here is someone, who want to follow this path.

I added the original recorded track to my dropbox. It’s just an extract, so it could be possible that you have to record a longer one.
You can find it here:
https://www.dropbox.com/s/xgwd0zw6dctgg15/scarletconsole.wav

Try reversing/backmasking the audio?

did anyone try to overlay the two sides? dunno how to do it myself…

The Kurt.2518:

khani.4786:

I was wondering if the black lines were hiding letters like the attached illusion….. but I haven’t had time to play with it myself. o.o

What black lines are you refering to? The spaces between the symbols or what? :P

Kind of like this….probably a dead end here too though….

I think I cracked the code. It’s weird – it has these same letters appearing over and over again:

BESURETODRINKYOUROVALTINE

Thedenofsin.7340:

I think I cracked the code. It’s weird – it has these same letters appearing over and over again:

BESURETODRINKYOUROVALTINE

BE SURE TO DRINK YOUR OVALTINE
that would be pretty random

For it to be valid though, you need to show how you got to it in order for us to replicate your result.

I Am Dansker.7105:

Thedenofsin.7340:

I think I cracked the code. It’s weird – it has these same letters appearing over and over again:

BESURETODRINKYOUROVALTINE

BE SURE TO DRINK YOUR OVALTINE
that would be pretty random

For it to be valid though, you need to show how you got to it in order for us to replicate your result.

That is called a troll.

On a more serious note I’ve started to do the frequency analysis but the approach is compronised. The boyz from reddit have done an amazing job at narrowing the set of symbols .

The problen is : how do you break the large symbols in smaller ones (I have a theory about it but there are more than one possible solution, see attachment)

Note : black line is most likely hypothesis ; green line is possible but unsure

VodCom.6924:

I Am Dansker.7105:

Thedenofsin.7340:

I think I cracked the code. It’s weird – it has these same letters appearing over and over again:

BESURETODRINKYOUROVALTINE

BE SURE TO DRINK YOUR OVALTINE
that would be pretty random

For it to be valid though, you need to show how you got to it in order for us to replicate your result.

That is called a troll.

It’s actually called a joke. I laughed because I’ve seen the movie. ^^

So, does anyone have suggestions of words to search for? Words with no repeating letters such as “Scarlet” are no use, words with only one repeating letter like “Primordus” are slightly better, and words with many repeating letters like “Mordremoth” are really good.

My current hypothesis is assuming total overlap. As in, EVERY “block” of 4 segments perfectly overlaps with the next block over, including dots. Based on the frequencies I’m 90% done with (I will edit this post to have them), I don’t think this is baseless at all. There are only 2 “nothing” blocks, and they occur in the first two lines of B and all of A thanks to the reflection aspect.

Link to frequencies

Based on this, there are 34 different symbols. If the only-dots ones are removed, it becomes 27. The x/y nomenclature is how many occurrences in B vs. how many in A.

WingedVictory.9013:

Luna.9640:

I’ll just leave this here :>

HAD to log on to say I’m glad I’m not the only one that sees Asura gates as Stargates >.> (Technically I call them Stargates too)

Technically Asuras are Asgardians actually :>

Hope you all actually find something. If it is just a wall to texture to make the screen look aesthetically suitable that’s an awful lot of effort for nothing. perhaps a Dev may add some insight. Though my gut feeling based on the presentation, repetition and limited nature of the ‘text’ is that it is just illusionary data feed text, akin to flashing machinery and excessive dials/readout in many ‘hi-tech’ game facilities (Rata Sum!).

TaCktiX.6729:

My current hypothesis is assuming total overlap. As in, EVERY “block” of 4 segments perfectly overlaps with the next block over, including dots. Based on the frequencies I’m 90% done with (I will edit this post to have them), I don’t think this is baseless at all. There are only 2 “nothing” blocks, and they occur in the first two lines of B and all of A thanks to the reflection aspect.

Link to frequencies

Based on this, there are 34 different symbols. If the only-dots ones are removed, it becomes 27. The x/y nomenclature is how many occurrences in B vs. how many in A.

That’s a lot of effort thank you. However a couple of remarks :

1) I don’t think that you splitted the symbols the right way look at the first image below : it appears that one cannot combine two symbols that overlap.

2) secondly it appears that there is some hidden rules on what symbols can be fused. My guesses :
-No overlapping (image 1)
- continuity of the line (image 2) or “T” shape (image 3) [an “L” after a “_” becomes a reversed “T”, however there are many possibilities that can give the same result]

I don’t know if these mean anything, but the icons for the Toxic Code Fragment and the Aetherblade Code Fragment have numbers on them.

………………. Toxic Code Fragment ……………….

Reads as: 01011001 or 010 100.

01011001 = Y in binary
010 100 = ??? Unknown

………………. Aetherblade Code Fragment ……………….

1010 = ??? Unknown

………………. Others ……………….

Includes: Watchwork Code Fragment, Molten Code Fragment, Scarlet’s Lockbox Code Fragment

The others may have something hidden in them as well. I’ll attach them. The Molten Code Fragment hnas some blue marking at the top right that I can’t identify. Perhaps an in game language like New Krytan?

The only thing that bothers is why the BD part is duplicated. For every AC. Can i ask you how did u name the parts ?

The thing that puzzles me is if letters can be combined by horizontal compression why is there a full gap between the ‘T’ and the backward ‘L’ near the middle of the first line of the 7 line block, since you could push those two symbols together with no loss of data?

Edit: Oh… Those vertical gaps may be the spaces between words. While ‘e’ is the most common letter in a typical blob of English, spaces are the most common character…

Nike.2631:

The thing that puzzles me is if letters can be combined by horizontal compression why is there a full gap between the ‘T’ and the backward ‘L’ near the middle of the first line of the 7 line block, since you could push those two symbols together with no loss of data?

Edit: Oh… Those vertical gaps may be the spaces between words. While ‘e’ is the most common letter in a typical blob of English, spaces are the most common character…

That is a good question albeit one that can be answered easily : it apears that there is no more than one point of junction betwen symbols. This rule is here to avoid confusion (I presume). Let us simulate your example (picture below).

I think it is quite impossible to solve the cipher the way we are going at it now. The problem are the merged symbols. Since they can be formed from multiple symbols like VodCom showed this already gives multiple translations.

However when assuming that symbols only can be merged if the line is equal for both symbols it is very hard to solve.
What I’ve been doing for quite some time is assigning every symbol to a letter so you can input the sequence in several programs.

The sequence I’m working with now is :
aeifjghfbofkl srfegtspgeaefp wfogpfsqdsvrf qigbfuphfjivcu uceghlhfuerocn gtcfxfovegsf yvudigpptahqoa jyeoceghlemfa pefafhnguhfeig

And am using http://www.blisstonia.com/software/WebDecrypto/
This program tries to find the right letters that result in sensible words. It has two options; 1 with known spaces and a option that finds spaces between words itself.
This last option seems the obvious one but it is very hard for the program to find solutions. You can test this by translating this sentence for example: Vjhqboi tmbb hiihjy Bmwrv Hqjd roui which is translatable when you use the “trust spaces” option but if you try it with the “find spaces” option after removing the spaces it will not make sense.
You can however give the program tips like v=s or roui=next but you will see that it needs a lot before it makes sense. It gets easies the longer the sequence is though.

But what my point is, you can see how difficult it is to simply try to assign the right letters to the code. However if you also can’t be sure about the symbols it will be downright impossible.

So either this is something the devs put in for fun, with never the intention of people solving it. Or we must restart and look for more secure ways to retrieve the code.

At first I started with a way to try to determine spaces between the words. I though there must be a reason why some symbols were stuck together. Maybe that were the actual words resulting in: aei f j ghf b o fkl srfeg t s p ge ae fpw f og p fs q ds v rfqig bkittenp hf ji vcuu ceghl hfu ero c ng t cf x f o vegs fy v u d ig p p tahqo aj yeo c e ghl e mfape fa fh ng u hfeig
But this is not right since too many words are too short.

So just to sum up. If we assume we are on the right path we need a way to determine spaces, else we need a whole new approach.

Okay so we know that the lines AC are the mirror image of BD. Maybe the first two lines of the text is a key to retrieve the real code. I can’t just accept that the developer responsible for creating the code was just lazy and copied/mirrored the first two lines of the texture to the bottom. There might be a meaning to why there are two copies of the text followed by AC.

What I did was copy paste the ‘key’ over each line. Before analyzing this I would get rid of the first two lines because it’s just AC+BD. Another thing that is a problem is that there are actually 5 lines of non-redundant code, but the key is 2 lines long. The second copy of the text is actually not retrieved by the key AC but by CA (because 2 does not fit into 5), presumably producing garbage code.

So if I am really onto something, the only lines of code that would make any sense are lines 3-7.

Edit: Right after i posted it, something struck my eye: If you put AC over BD you get an obvious mirror pattern. But if you put CA over BD you can see that C+B = mirrored(A+D), which seems non-trivial to me. You can see it in the picture: lines 8&9.

Each symbol with a point next to it is visible multiple times. I connected each of them to their corresponding symbol, and ALL of the connections were either linear or diagonal. As you can see on the picture, there`s a pattern. They are perfectly linear and or diagonal, but on the picture, they seem off, but if you put time behind it, they are linear.

I haven`t tried matching the first pattern with the second code, though. We might see a message or a picture, since the code changed, like shown in the OP`s picture.

Rabe.2456:

Edit: Right after i posted it, something struck my eye: If you put AC over BD you get an obvious mirror pattern. But if you put CA over BD you can see that C+B = mirrored(A+D), which seems non-trivial to me. You can see it in the picture: lines 8&9.

That’s predictable unfortunately. B = A backwards and C = D backwards. Therefore superimposing B and C is superimposing A backwards and D backwards.

Kioshi.6597:

The sequence I’m working with now is :
aeifjghfbofkl srfegtspgeaefp wfogpfsqdsvrf qigbfuphfjivcu uceghlhfuerocn gtcfxfovegsf yvudigpptahqoa jyeoceghlemfa pefafhnguhfeig

Try doing away with the last 2 lines of text. The symbols in the two line phrase are the same as the symbols in the first two lines of the seven line phrase, but backwards. Because of this I would suggest entering data from the seven line phrase backwards.

VodCom.6924:

That is a good question albeit one that can be answered easily : it apears that there is no more than one point of junction betwen symbols. This rule is here to avoid confusion (I presume). Let us simulate your example (picture below).

Except to form that box and cause a loss of data you moved the two clusters towards each other twice. A single shift would place the tip of edge of the T above the tip of the tail of the reverse-L without distorting or obscuring either formation.

That suggest the compression is either erratic (encoder wasn’t paying attention), purely aesthetic (some compressions are skipped so that the entire line is ‘justified’ to a target width), or its absence means something.

Looking at other compressible empty gaps, those could be spaces in a short phrase. I’m not convinced I’m on to something, but I’ll keep pecking at it .

I think the forward and reversed text might be an artifact of Scarlet’s divided psyche.

Maybe its a signature and ‘both of them’ signed it. It will crack me up if ‘Scarlet Briar’ and ‘Mordremoth’ are mirror images when viewed through her personal cypher .

As time passes I am more and more convinced that this code is an hoax created for purely aesthetic reasons. Here are my reasons :

1) Some lines are mirrored, others are not : cheap process to generate graphically different lines of symbols that can be used to fill up Scarlet’s screen.

2) No obvious way to disentangle large symbols. Even assuming that there is 1 point of junction between two fused symbols, there are many possibilities for some symbols.

3) Even if we manage to disentangle our symbols the right way, the text resists frequential analysis (especially the analysis of bigrams). This is because of mirrored lines that screw up frequencies.

Conclusion : I am not investing more time to decipher something that is not meant to be cracked.

VodCom.6924:

As time passes I am more and more convinced that this code is an hoax created for purely aesthetic reasons. Here are my reasons :

1) Some lines are mirrored, others are not : cheap process to generate graphically different lines of symbols that can be used to fill up Scarlet’s screen.

2) No obvious way to disentangle large symbols. Even assuming that there is 1 point of junction between two fused symbols, there are many possibilities for some symbols.

3) Even if we manage to disentangle our symbols the right way, the text resists frequential analysis (especially the analysis of bigrams). This is because of mirrorred lines that screw up frequencies.

Conclusion : I am not investing more time to decipher something that is not meant to be cracked.

I agree. I’ve come to the same conclusion. Mainly because of point 2.

Hope once the Living Story arch is over the devs can tell us if it actually was relevant

I’ve been reading this thread and started attempting to crack the code ( if it even is a code) after several permutations of a Pigpen-like cipher the only headway i made was erronously decoding the first line into “ICYGVGEADDAQ_”.

which of course could be Icy GvG & add a Q… maybe someone can use that.

i was thinking the last word could be quest which made me want to continue decoding to see what kind of quest it would of been but i got lost realizing i had messed up in the decoing somewhere.

Cormac.3871:

Rabe.2456:

Edit: Right after i posted it, something struck my eye: If you put AC over BD you get an obvious mirror pattern. But if you put CA over BD you can see that C+B = mirrored(A+D), which seems non-trivial to me. You can see it in the picture: lines 8&9.

That’s predictable unfortunately. B = A backwards and C = D backwards. Therefore superimposing B and C is superimposing A backwards and D backwards.

Hm, sadly you are right.

I wanted to point out, to all those who think the last section of the code may be irrelevant, that if read from left to right the last section is a completely different set of symbols and that just taking it as the inverse may not be the proper solution.

A couple of things that strike me:

1. There an actual ‘texture’ pulled from the .dat file on the reddit discussion. Using the actual texture, as opposed to a snapshot of this texture moving in an animation in Scarlets lab, would make trying to translate it a lot easier. http://imgur.com/XJojIcg

2. On this texture the following can be concluded. It is a texture that consists of 2 blocks of 7 lines, which are the same. And 2 lines at the bottom which are the first 2 lines of each block ‘mirrored’.

2a. This means that if there is a text hidden here, it is repeated 2×. And we would only have to point our efforts to a block of 7 lines.

2b. If there is a text here, there would be no reason to ‘mirror’ the first two lines at the bottom. Unless it is a clue of sorts.

3. The lines are evenly spaced horizontally, and there is no way to identify a vertical spacing. So the text is to be read from left to right (or right to left). And not in vertical lines from top to bottom (or bottom to top).

3a. A problem that arises to decipher this text is how to determine how ‘wide’ one character is.

4. Seeing all known Tyrian languages are ciphers themselves, looking for a double cipher isn’t needed. One could take into consideration though that ‘Asurran’ language has the same symbol for C, K & Q . Seeing Scarlet studied a lot with them, it may help taking this into consideration.

- So far only thing that I did was to mirror the entire texture horizontally. And from that it seems a tad more clear as to where spaces ‘may’ lie in the actual text. for some reason the image flipped horizontally seems to make more sense than not. This obviously entirely based upon some ‘feel’ of what makes sense :P … then again, if there is something in this texture, then mirroring the last 2 sentences wouldn’t make any sense unless they were in there as a clue.

- Another thing I notice is that there are several symbols that are a ‘vertical’ mirror of each other. Mainly the T but also the ? on it’s side. Together with all the other symbols that had people suggest a certain ‘block’ cypher. Which is just another way to make a substitution cypher, might well give hints as to how the cypher is constructed.

As there are a lot of ’ ’ and L U that are ‘mirrored’ vertically or horizontally. or ‘on their sides’. Still though, correlation with structure and logic does not mean causality…

As there also appear to be a couple of ‘bigger’ symbols, of which it is unclear if they are a symbol on their own, or consist of ‘merged’ loose symbols. Which basically brings me back to the issue with ‘how wide’ these symbols are..

Yeah, so far the hardest thing in trying to decipher this are the combinations of symbols that ‘when close together’ seem to form another possible symbol. This is best illustrated by looking at the ‘proposed’ symbols on Reddit: http://imgur.com/KzO6RO3

And see how the (1st line 3rd symbol) and (2nd line 4th symbol) make the (2nd line 1st symbol). And there are various other combinations possible too. For example the = and the | symbols can create the C symbols quite easily. So without knowing how to raster these or divide them up, deciphering this (if there even is a translation) would be incredible hard.

Anahiem.2819:

I’ve been reading this thread and started attempting to crack the code ( if it even is a code) after several permutations of a Pigpen-like cipher the only headway i made was erronously decoding the first line into “ICYGVGEADDAQ_”.

which of course could be Icy GvG & add a Q… maybe someone can use that.

i was thinking the last word could be quest which made me want to continue decoding to see what kind of quest it would of been but i got lost realizing i had messed up in the decoding somewhere.

i should add, that i was also using the texture pulled from the game files found on reddit to deciper the text. i made the assumption that each “square” represented one character and that there were no spaces. i think in all my attempt i found a total of 43 possible symbols that could be made using the lines and dots arranged in the “square”.

i had many different versions of a cipher drawn out containing varying numbers of symbols but when i got to applying the ciphers to the string of symbols with rules i assumed about them, my ciphers didnt hold up. there always seemed to be a few symbols my ciphers didnt include. i eventually gave up due to the fact that there is no confirmation that it is even a code at all. If a Dev steps out and confirm it as being a hidden message, i’ll head right back to it but until then i’ll just leave it to someone else.

One last thing, if anyone determines a proper pattern to the symbols, converting them to letters, and doesnt see anything coherent they should try using OMADD as the cipher key. My thought was that it was super encrypted but i seemed to make more progress just doing a symbol= # = Letter decryption.

If you assume every square represents one symbol, then you can’t tell to which squares the vertical lines and points belong to. You’d have to make a rule to ensure uniqueness. For example if you say every square can only have a vertical line on the left side, you’d be okay. But that also limits the number of possible symbols.

Rabe.2456:

If you assume every square represents one symbol, then you can’t tell to which squares the vertical lines and points belong to. You’d have to make a rule to ensure uniqueness. For example if you say every square can only have a vertical line on the left side, you’d be okay. But that also limits the number of possible symbols.

I’ve been trying to do this for quite some time, but there are symbols that cannot be properly separated. See attachments in earlier posts in the thread.

You can seperate them. Take that example rule to always take the left symbols of each block. The seperation is absolutely unique and consistent, but it’s somewhat arbitrary.

Edit: This example fails if there is a vertical stroke at the end of a line (which is the case, for example line2, line3, …).

Rabe.2456:

You can seperate them. Take that example rule to always take the left symbols of each block. The seperation is absolutely unique and consistent, but it’s somewhat arbitrary.

Edit: This example fails if there is a vertical stroke at the end of a line (which is the case, for example line2, line3, …).

That rule doesn’t work. Look at the images attached in this post :::
https://forum-en.gw2archive.eu/forum/livingworld/madness/Cracking-Code-in-Scarlet-s-Lair-SPOILERS/3598370

Half Life 3 – I mean . . . expansion confirmed.

I still think my “everything overlaps” approach is the one most likely to be correct, since “left only” and “right only” approaches run into the fact that both sides generally have something on either edge.