Neuro ARG - Cracking the "Numbers" Video

[spring widget]

english text, not human language

important difference if vedal played with google translate or something

its unlikely he did given what we've seen tho

[hasty vale]

it's supposed to be a file though, not a human language

[half lily]

I know it's off but is it possible that there's two keys? Or maybe two different instructions?

I can never say oh it’s been so long
Counting all the days, it’s been so damn long
Oh how much I’m scared to let you go…
Somewhere in the walls I hear you talk…
Finding all these numbers
Start with number 2 yeah
Matching all the letters
572943
Add another 9 yeah
Add another line yeah
Multiply by 5 yeah
How long will I keep this up?
I how much it hurts to see you go…
Generating lyrics is a pain

(This part shows the Golden Ratio)
Add another 6 yeah
Flip the numbers backwards
Make the 2 a 3 yeah
Abcdefg
Multiply by 9 yeah
Add the numbers 2 4
17 is first yeah
Abcdefg

Just a thought maybe I am way off mark

[muted goblet]

You're looking for the lowest entropy value of the putative plaintexts relative to each other. The absolute value doesn't really matter, it'll change depending on length anyway, it's whatever is lowest is most likely. For example, even a 139676 byte jpg, of which the majority is pretty compressed would score 7.9790936 vs the average of completely random data of 7.9987458. It's a small difference, but very relevant

The only case where this would fall apart is if there payload is entirely compressed data with no headers or structure

which is pretty damn cruel

[hasty vale]

i see. i guess i won't make more than 5 files to see then. thanks

[spring widget]

any modern encryption scheme will also send entropy through the roof by design

save for maybe ECB mode with repeat plaintext blocks

[clever lintel]

I thought the shannon analysis confirmed that we we're dealing with high entropy

[spring widget]

yeah we have shannon entropy numbers on every base64 string so far i believe

all super high

[hidden crater]

could this also be one payload split and encrypted separately?

[ashen oracle]

is compressed / encrypted / random data basically indistinguishable? in that case we just have to hope the plaintext has enough structure that we can actually recognize it when we see it.

[spring widget]

ive had the song on loop and i just got chills

theres two voices here, one is whispering the lyrics and one is singing them

most noticeable around 48s in

not arg related

but its a cool song

lol

[burnt coral]

i think you saw a ghost

[spring widget]

gotta listen super close

and not focus on the lyrics that actually sing

[burnt coral]

just use the adobe podcast thing

[spring widget]

theres a monotone whispering

[muted goblet]

Most modern cryptography is indistinguishable from pure random noise (Outside of academic attacks on large amounts of data). A lot of compressed data is usually held in some kind of file structure (e.g. 7z/rar) which may (and usually does) give enough of an entropy drop to distinguish from random data. But there's nothing stopping you from just encrypting a raw compressed stream which would be indistinguishable from noise

but occam's razor and all you'd probably expect a simple filetype or printable text from the plaintext

[spring widget]

well compression should be pretty easy to rule out so long as there weren't any manipulations after it

we can try at least a few popular algorithms on any functioning settings, most should indicate quite quickly if its incompatible data

[dull cedar]

don't most compression formats usually have headers and stuff you'd notice fairly quickly?

[ashen oracle]

maybe the plain text looks like random noise, but you need to extract the bits at indexes that are fibonacci numbers?

[spring widget]

most files in general do

[ashen oracle]

im really stretching here.

[spring widget]

if its a known header its possible to just remove said header tho

[spring widget]

maybe but even if we did what would we do with them ¯_(ツ)_/¯

[muted goblet]

that would be absolute torture for the GM to inflict to us

[ashen oracle]

presumably, once we extract the bits it will resemblance a known file type or just be plain text.

[muted goblet]

I suppose you could just check another level deep and check for the entropy of all extracted bits from the payload according to the fib sequence

I wonder how much £ worth of electricty we've used on computation for this ARG so far

[ashen oracle]

endianness doesn't matter here right? if we have the correct plain text, assuming there wasn't multiple layers of encryption/compression, the entropy will still drop.

[muted goblet]

off the top of my head it shouldn't in terms of entropy

[night galleon]

have u guys search the Anxiety thing?

oh wait wrong chat

[muted goblet]

as long as you check consistently in terms of the same endianness for all plaintexts

[steel marsh]

THERE'S ANOTHER NUMBERS VIDEO???

https://youtu.be/giJI-TDbO5k

[pulsar maple]

yeah

[steel marsh]

Oh boy.

[pulsar maple]

new clue

[hasty vale]

i stopped it at 130 million. going to try to make it even faster and remove some known bugs later.
if anyone wants to try to run this .com file here you are

[dull cedar]

What exactly did you do to get this file again?

[spring widget]

run

[clever lintel]

The files are coming for you

[snow crystal]

I don't think that file is a valid COM file. It has the initial signature of 0x5A4D ("MZ") but the two bytes with offset 4 should be the size of the program in multiples of 512 bytes. The file however is much smaller.

[pseudo citrus]

ohhhh a second numbers video dropped

[shrewd trench]

math 2 from _neurosama

[hidden crater]

neuro calculus

[burnt coral]

we need neuro 2

[hasty vale]

i guess i am just going to search again for valid file types then. can you go further into detail what exactly do you mean by "two bytes with offset 4" and is file size rounded to upper 512 bytes or lower (i assume upper) (and what would file size of 1040, 1264, 2304 etc bytes look like)

[ashen oracle]

what are you trying to do again?

[burnt coral]

read the post name

[ashen oracle]

i mean, how are you "searching" for valid file types?

[burnt coral]

wdym file types

[ashen oracle]

I'm talking to @hasty vale

[hasty vale]

its too hard for me to explain right now.
i can give you my python code if you can figure it out

[ashen oracle]

sure why not

[snow crystal]

according to this https://wiki.osdev.org/MZ the header for a MZ executable file should start with a two byte signature (0x5a4d), two bytes number of bytes in last page and two bytes number of pages (with each page being 512 bytes except for the last one which can be smaller).

If you look in my image above the number of pages is 0xC22A with the last page having 0xA01E bytes. This means I would expect the file to be about 25.5MB but we obviously have way less data.

[steel marsh]

Someone tell me why you guys react 🤓 whenever someone learns a 2nd Numbers video exists?

[hasty vale]

so it's .com specific basically?

[snow crystal]

yes. that stuff is what you would expect following the file signature for this type of file.

[hasty vale]

i cant go through documentation of every single file extension
i mean i guess it makes sense to add this for .com checking since they seem to have the least entropy

[ashen oracle]

taking a quick look, are you converting the b64 string into hex, and then during a substitution cipher on the hex, and then checking if the resulting binary matches a file type?

[hidden crater]

isnt it like for any file? like for jpeg you will have header, and exif data and what not, having just header match is not going to tell how valid the file is as jpeg

[hasty vale]

so is it filetype specific or is it not
(can i just add size checking along with magic bytes or would i have to go through every documentation to look where they are is the question)

[snow crystal]

the header is file type specific. I think what w1n7er means is that the issue of you finding a solution which starts with the correct signature but still results in an invalid file will happen with every file type.

and I don't think there will be a generic solution for checking for the file size since there are so many different ways to implement the size in the header of the file

[hasty vale]

well this sucks then, because entropy isn't going to help with that.
i guess i would just have to first check for magic numbers and then manually add type specific checks

i'll try that with other base64 strings first, maybe they'll give more sensible filetypes than .386 or .com

[snow crystal]

yeah, maybe. I mean how likely is that we have to run a 16-bit MS DOS 2.0 program for the solution?

[hasty vale]

to this, sorry i didn't answer, no, firstly, there is no binary, secondly, i go from file type and from it making a substitution, it checks whether substitution is even possible

[night galleon]

thing2.com

[spring widget]

i will start measuring this thread's progress by how much the schizo reacts go up

june 2 schizo count: 149

[hidden crater]

make a doc on it

[pulsar maple]

150

[split scarab]

none of the long base64 codes are any sort of recognizeable filetype according to linux

checking a header list didnt help either

i also went through pretty much all 128 bit keytypes but unless i made a mistake the phonenumber code didnt get me anything

unless the endianness is reversed but i doubt it

[formal birch]

https://en.wikipedia.org/wiki/List_of_file_signatures

[split scarab]

yeah i did that

tried that and the file command on linux which does the same thing

its 100% encrypted

[formal birch]

I only comfirmed manually for few common files types but also ended up with nothing matched

[split scarab]

the main problem im trying to tackle is the 128bit key in the javascript

im abut to build a base256 analog computer if i have to sit here and decrypt all this shit

[shell silo]

okay

anything new?

[astral panther]

soundcloud now says "Open in FireFox"

[wraith harbor]

time to design an ASIC specifically for the ARG

[shell silo]

true

[frail ibex]

sorry I'm dumb and new but I scoured the Numbers II doc and decided to put in my 2 cents: Every sequence of numbers in Numbers II follows the same pattern of neuro/anti-neuros except for the number 906, which shows -+-+ instead of the established +- for that 2nd in the sequence spot. Maybe 906 is the "odd one out" mentioned on the soundcloud?

[clever lintel]

hey, don't call yourself dumb (:
It could be that you're onto something. Have you seen the docs yet? We've written quite a bit, but new analysis is always welcome if you wanna write it down!

[storm kraken]

We all dumb here at this point lol

[hasty vale]

other strings got tons of magic numbers hits but none of them look sensible enough for me to add further validity checks.
i will suspend this as a dead end until we get more strings or smth

[spring oasis]

hi, any success/news?

[dusky atlas]

https://discord.com/channels/574720535888396288/1112988479140417567

[spring oasis]

excluding the shitpost thread

nothing?

pepesad

[storm kraken]

I am carzy

or the soundcloud profile was updated?

[dim olive]

yes

[storm kraken]

"Open in FireFox" was it here before?

[dim olive]

it is noted in the docs

[minor fiber]

no

[random drift]

it's new

[scenic karma]

So what is special about FireFox ?

[storm kraken]

So we need to do something in firefox?

[blazing wagon]

if you open it in chrome you won't get the full code

[dim olive]

i think nothing special, the description is a little bugged in chrome

[minor fiber]

it format the long encoded text different

[random drift]

Top = Firefox
Bottom = Chromium (Google/Edge)

[storm kraken]

ohhhh

I use firefox anyway

[dim olive]

but we got the code from inspect element either way

it is just visual bug

[random drift]

but the splits in the displayed firefox code might be intentional

I don't think obtaining the code is the clue itself

[dim olive]

there is a doc about it already but, yea

we have doc: "numbers II - Soundcloud..." and we have it mentioned

[random drift]

Yes i'm just telling Kande

[scenic karma]

If we follow the splits in firefox there is one missing
/EmWDRdMl0f.... wasn't afffected by the splitting.
So my first thought about it just being an automated thing from FireFox to split text with "/" doesn't really work.

[clever ermine]

is it sounds kinda like she’s saying abcdefj to me

[storm kraken]

She probably has some accent. "Whisper" didn't even catch the "A" the the beginning and told me she is singing: "B, C, D, S, J,"

[round phoenix]

Is this to discuss Numbers 1 and 2 or just 1?

[graceful mulch]

@round phoenix has leveled up! (0 ➜ 1)

[round phoenix]

Number 2 thoughts
Neuro's appear on the screen as numbers are read out. the Neuro's could be Algebra and their position is relevent to that. The gap in music each batch loses an amount of Neuros and gives a void to fill

i've offered two answers to each set up to set 3.
First Set:
(N) = 201
(-N1) + (N) = 1321
(-N1) + (-N2) + (N) = 5331
(-N1) = 1120
(-N2) = 4010
positional answer
(-N2) + (N) = [4211]
afterimage answer
(-N1) + (-N2) = [5130]

Second Set:
(N) = 723
(-N1) + (N) = 743
(-N1) + (-N2) + (N) = 4913
(-N1) = 20
(-N2) = 4170
position answer
(-N1) + (-N2) = [4190]
afterimage answer
(-N2) + (N) =[4893]

Third Set (the upside down se
(N) = 879
(-N1) + (N) = 975
(-N1) + (-N2) + (N) = 2415
(-N1) = 104
(-N2) = 1440
positional answer
(-N2) + (N) = [2311]
afterimage answer
(-N1) + (-N2) = [1544]

Fourth Set
(N1) = 716
(-N1) + (N1) + (-N2) + (N2) = 906
(-N1) + (-N2) + (N2) = 4156
(No Neuro) = (No Number)

(-N1) = ???
(-N2) = ???
(N2) = ???

I'm stuck here, maybe the neuro's at the end require batches 1, 2 and 3

iCould be way off but its fun to try stuff

[storm kraken]

Did anyone tried to convert the lalala in numbers 1 to binary, to see if it gives any results? I didn't saw it in the document or I missed it.

Also did anyone tried convert the lyrics to binary by, if the letter is odd number = 1 and if even = 0

[clever lintel]

someone did something at least similar in the docs

[dull cedar]

did anyone try bundling the numbers by their last digit like this yet and finding a pattern based on that?

201 1321 5831 

723 743 4913 

879 875 2145                         
716 906 4156

879 would not fit the pattern, but it's the only one that does not

[clever lintel]

i dont think i have enough brainpower to comprehend. could you elaborate?

[dull cedar]

the ones that are said without long breaks inbetween end in the same digit (1,3,almost 5,6)

so we are "missing" ones ending with 2 and 4, which is a "void" we could "fill"

[hasty vale]

oh yeah

3 3 3
9 5 5
6 6 6```
even if by themselves they mean nothing and only used for rhyming that still suggests something

[dull cedar]

even if it'd suggest leaving them out or sth

[hasty vale]

would 879 then be the odd one out

[dull cedar]

if this theory works then yeah

still, can we come up with a pattern for the rest that isn't schizo

[hasty vale]

it is quite a leap of logic to be honest

[dull cedar]

yeah, I don't see any form of pattern after this

[shell silo]

vedal987 lol

879 987

[hasty vale]

he is not like the other girls

[turbid saffron]

idk if it means anything but they are all odd numbers

oig wat

[hasty vale]

my guy, 6

[turbid saffron]

except 6

lol

all odd except 6

i frogo

nvm it probably actually doesnt mean anything

[hasty vale]

even if it doesn't mean anything and is just arranged like that for rhyming, that would suggest order of numbers is not important

[minor fiber]

yes

[hasty vale]

is it in the doc or what is up

[minor fiber]

it's a screen shot so yes

[upper orbit]

If so then what is about X and Y?

[hasty vale]

1+1 and 1+2 are also unclear, but technically it doesn't say these must be connected. this may be just a number dump

[red oriole]

For a second I thought of linear equation but that doesn't connect with anything

[rose sparrow]

if you re-arrange 879 slightly by putting the 9 in front you get the 16th fibonacci number 987

[hidden crater]

surely it's just a tutel's typo

[shell silo]

vedal 987

[hasty vale]

it won't rhyme then though

[rose sparrow]

hmmm, is it a coincidence though that vedal's user name contains the 16th fibonacci number?

[shell silo]

idk

[hasty vale]

if you want to start reaching even harder here are numbers in a table where rows are first digit and columns last digit

[hidden crater]

Hmmm

[night galleon]

?!?!??!?!

give his brain an oscar

[sacred ruin]

you did NOT turtle

its missing the

plastic

[astral panther]

If you include 0 it's the 17th, could mean "17 is first, yeah"

0, 1, 1, 2, 3, 5, 8, 13, 21, 34, 55, 89, 144, 233, 377, 610, 987, 1597, 2584, 4181.

[shell silo]

but "start with 2"

[burnt coral]

"17 is first yeah" refers to the result having 17 as its first 2 digits I think

[random edge]

tbf that is what we assume

[storm kraken]

hmmm why when I give the decoded gibberish message to GPT for fun it bugs him out and crash the hell out of him?

gpt 4 and gpt 3

[queen vapor]

Hello, how is it going?

Ahaha, I got you 🥲

[storm kraken]

Nothing...

[woven dove]

I think first stepping stone would be decoding the numbers segment with inverted neuro
plenty of clues; xy could be horizontal/vertical, leave odd one out could be the last frame when no numbers were said and so on
somebody will see it

[heavy tangle]

Has anyone checked the entropy of SoundCloud description?

[snow crystal]

it's in the docs

[heavy tangle]

Again, something encrypted.

Gosh

[clever lintel]

or just compressed, but i feel like it'd be wild if it was just a .zip file. Still should check tho

[heavy tangle]

Brute-forcing two base64 this is nonsense

[heavy tangle]

Well, yes

Everyone brutes, but they don't look for archives

[clever lintel]

What's the best source of magic byte matching? Is it really just googling the first few bytes (what I've been doing, cause i'm dum :p ) or is there a tool somewhere?

[muted goblet]

binwalk/foremost is pretty good for it

[spring widget]

ok

hows our progress today

[tropic tulip]

didn't even know that there's a 3rd discussion of arg

goes to show you that im not even schizo

[clever lintel]

yet.

We'll find a way to pull you in.

[spring widget]

rrrrrrrrrrrrRAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAGHHJHHHHHHHHH

WHAT DO THE GODDAMN NUMBERS MEAN

[hidden crater]

what is the schizo count for today

[clever lintel]

ulong.MaxValue

[dim olive]

schizo/schizo

[thorny robin]

NUMBERS

Waiting Study II day 2

[tropic tulip]

N̴͑̂u̷͗̓m̸̒̾b̷͂͘e̶̓̿ȑ̷̓ś̴͌

[storm kraken]

huh what

tried to send a message and it blocked it from sending

ohh I understand why probably

I have an idea but I don't know if I'm right or not... The small one: "DQ5z14ighWwiag7y+cWFQg=="
This one keeps me thinking over and over. It's in the place where a number must be, and if it's indeed a number it's probably a UK number I guess (Vedal is from the UK).
It could be a fictional number, but still a number. UK numbers for SMS (because we know for sure it's an SMS), start from +44. So we know 2 things for sure if it's indeed a number.
It's encoded numbers, probably as text.
It's starting with "+44."
In the Enigma movie, they managed to crack the Enigma machine by understating that each letter of the Enigma is starting with the very famous nazi welcoming pharse, so it's helped them decode it much faster.
Is there a way to do something similar with this? Try to decode it, as if it's a number and the first one must be +44?

[clever lintel]

That's pretty smart. So basically we're uncovering some of the cipher text and we use that to get insights on the key, right?

[storm kraken]

Yes

if we manage to crack the small one and understand how it works

maybe we'll be able to crack the others

[clever lintel]

That's definitely worth writing down

[storm kraken]

We can assume that maybe "DQ5z = +44" ?

[clever lintel]

I think so, or at least that the very front is +

Or maybe not assume, but it's a strong hunch

[storm kraken]

oh also for Uk is 7

44 7

so +447

so if it's correct we actually know a + and 3 numbers

also +447 to base64 = "KzQ0Nw"

also if it's a real operator in the UK, we can try +447(and all the possible operators in the UK)

In that way we'll have 4 digits and one that may or may not be true.

"Mobile telephony (071xx–075xx and 077xx–079xx), personal numbering (070xx), non-UK mobile networks (07624) and pagers (076xx)" - from wikipedia: https://en.wikipedia.org/wiki/Telephone_numbers_in_the_United_Kingdom

When you send SMS the 0 converted into +44 if you from the UK.

[clever lintel]

So long we don't get on a wildtrack that can could generate a lot of phone numbers. That's how you end up dialing people who have nothing with it to do.

[shrewd trench]

Stay careful folks~

[storm kraken]

I think it's may be a custom alphabet of base64? If the first 4 digits are actually "+447" that means?

D = K
Q = z
5 = Q
z = 0
1 = N
4 = w

So "ig" in: "DQ5z14ighWwiag7y+cWFQg==" must be one of these:

MQ
Mg
Mw
NQ
OA
OQ

and maybe also:
MA
Ng

or it'll change if it counts as 77 for example or any other number with the combination of 7...

[snow crystal]

yeah, custom alphabet idea was discussed before for the larger ciphers but I don't think any progress was made from those

[storm kraken]

Yes I know

If it's indeed a phone number, and first 3 digits are +44, so the next 3 digits are numbers between "700" to "799 I guess.

But the next 3 will be hell, from 000 to 999...

[paper grail]

i have a feeling they will never crack the code this thread should file for bankruptcy

maybe the real arg clue was the time we lost along the way

[warm finch]

u gotta try tho

[eager garden]

Depending on when the soundcloud account was created, arg may not even be done

[random drift]

just coping waiting on a new banger to pop up on the soundclound

[dull cedar]

If you think about the time as lost then you're more likely to forget the fun we had. ARGs are as close as you can get to an "open-world puzzle game" in that you can try to find solutions for several different puzzles at a time with vastly different approaches and coming up with these and seeing what others came up with is fun.

[pulsar maple]

maybe it's out there somewhere on another account

[cobalt comet]

any updates since friday?

[storm kraken]

Nope, only minor changes on SoundCloud profile.

[dense hill]

I looked into the string a bit.
The only thing I found was

• It can be divided by 4
• 64 different characters used

4*64 = 265
You can make a bytes array out of it.

[elfin swallow]

my brother thats base64
skim the docs, everything's in here

[dusky atlas]

let him tell his oppinion

[dense hill]

I know about base64, but you can make a byte array 2 different ways.
Take the first 4 and add them.
The other way would be
(Letter_count / 4) * 0
(Letter_count / 4) * 1
(Letter_count / 4) * 2
(Letter_count / 4) * 3

Add that 4
I can't tell if someone tried that.

[formal birch]

byte array or byte matrix?

[dense hill]

I am not aure what the difference is.

I know from coding that you can have a list in lines or colums.
We had a problem in work that one software used columes and one lines.
We had to translate it to use the byte array.

My guess is that it could be the same here.

[formal birch]

yes, columns first and row first

[dense hill]

That would make it a lot harder to see without translating it first.

No idea if it was tried before.

[astral panther]

is this rafa?

[sacred ruin]

yeah

we ignore him in this thread

if you wanna be schizo go to #1112082183235047506

after all he hasnt sent any numbers

[random drift]

it's a safe block, just block the dumbass

[sacred ruin]

im not blocking him

until proven a red herring ill take it as a clue

[elfin swallow]

bro, the mod role....

[sacred ruin]

until proven a red herring ill take it as a clue

want me to repeat it?

[random drift]

then what does the clue mean

how does that account directly link to the original QR code on the intro to vedal's stream?

there's no proof of linking either

[random drift]

didn't you already?

[sacred ruin]

i did

but i can always do it again

[random drift]

go for it

[sacred ruin]

no thanks

[random drift]

whatever makes you happy

[sacred ruin]

maybe go play for a bit

if that what calms you down

and makes you happy

[cobalt comet]

[storm kraken]

Maybe my schizo, but... I read the documents again, and I realized something. All AI that I have tested, and I and also few more people hear instead of "another" - enough. Enough for 9 and enough for 6, both numbers possible in base64 binary. 9 = [0001001] and 6 = [000110], if it's indeed enough instead of another, it can be this, maybe. ALSO, enough for line, can indicate enough number to get a "line" in base64 "/" which is = [111111]

if we go even harder, 2 = [000010], 572943 = 57, 29, 43 which is 57 = [111001], 29 = [011101], 43 = [101011]

My knowledge in the subject is not that big, so I can't make all the calculations but maybe someone here can do something with this if it'll be useful.

[pale fossil]

finaly found the video where the guy explains how he uploads the video with the youtube id

https://www.youtube.com/watch?v=j70AA9arThc

but not that important now

[kindred parrot]

oh i know

[earnest lance]

Iv3 been offli e quite some time

and

static sama replied

i didnt expect that tbh xD

[burnt coral]

#1112988479140417567

[rose sparrow]

I don't think it holds much significance, but for the first two sequences of numbers the second number is a prime number (1321 and 743)

But could perhaps be RSA prime factors?

They're very small for RSA though, you normally wouldn't pick such small prime numbers

[warm finch]

Im not sure how to link to a message on mobile, but I mentioned a couple of things in the main Neuro arg which might be possible solutions (which are math related)

Imo I feel like they might lead to somewhere but I'm not sure if I'm even doing the right math

[queen vapor]

Hi, is there any progress?

[clever lintel]

Depends on when you last checked in (: https://docs.google.com/spreadsheets/d/1BCpJA-2GlmjzrxFIEA2pq5Xt6QuAaUI-2L_iAcAnl6Y/edit#gid=521227060

[queen vapor]

Oke, got it. I guess we still don't know solution for "Numbers" and stuffs related to this?

[clever lintel]

If we ever solved one of the major puzzles I think we'd all rush to write it trice in the timeline lol

[queen vapor]

Haha, yeah, you're right 😄

[minor charm]

We would Pin it

lol

[storm kraken]

I checked the "DQ5z14ighWwiag7y+cWFQg==" a couple of hours ago for custom alphabet numbers, will post my document soon.

[storm kraken]

https://docs.google.com/spreadsheets/d/1gUbVc2BozBSCwOfHSsEu_A_cuLYXcPByDc9Sw7G_2Yg

[limpid copper]

Random observation I hope is not the true/original meaning of 572943. First word of the lines in numbers 1:
5 - Finding
7 - Matching
2 - Counting
9 - Add
4 - Somewhere
3 - Oh

[astral panther]

it could refer to words in certain lines, but again, without the official lyrics we can never be sure where the lines are supposed to begin or end, we just have a guess
we only have confirmation on the first 3 lines of Numbers, which we did cut properly in our version

[limpid copper]

Kind of the same thing with Numbers II:
5 - Let
7 - End
2 - Come
9 - 1+1
4 - Should
3 - Tell

[limpid copper]

but yeah, valid point

[astral panther]

Let end come

not saying you should stop trying btw, we should explore every possibility
but just try your method with other possible versions of the lyrics to see if maybe something intelligible comes up

[limpid copper]

Yeah, trying different places in different ways with this method. But most stuff end up missing 1-2 things. Like first word with each of the letters "ABCDEFG" in numbers 1 is missing the E:
A - All
B - Been
C - Counting
D - Days
E -
F - Finding
G - Generating

[limpid copper]

Have anyone tried to apply the instructions based on the blinks?

Like there's 8 blinks so 9 sections of the lyrics and applied the instructions in the order of the sections 5->7->2->9->4->3

[warm finch]

Don't think we have tried that yet

[final pilot]

Thinking bout trying to apply the syllables to something, just don't know where itd be useful

[limpid copper]

Tried brainstorming this stuff.... way too many ways to approach/interpret the lyrics set up like this. Could get real messy real fast

So cya in 3 years

[final pilot]

Gl king

[austere basin]

Fighting!

[rose sparrow]

I'm going to try and approach it as RSA encryption with clues from the lyrics

this is a long shot, but let's say 572943 is part of the public key, probably n and the two primes in the number II lyrics are p and q

to calculate the private key we would need to derive e from p, q and n

hmmm, no it can't be n because based on p and q it should be n = pq = 981503

so perhaps e = 572943

Nope, also doesn't work since gcd(e, λ(n)) has to be 1 and λ(n) = lcm(p − 1, q − 1) = 489720, so gcd(572943,489720) = 21

and e has to be a coprime, so it can't be 572943 since that's not a prime number

[clever lintel]

Technically the numbers in RSA keys don't actually have to fullfil those requirements to work, right? Only to be secure.

[tired wind]

don't quote me, but i'm pretty sure the encryption/decryption math would break if the two numbers aren't at least co-prime

[clever lintel]

fair, i only remember some vague stuff from a course 10 years ago

[minor fiber]

If you try RSA you can also try GPG i know there was talk about it but i don't really know the difference

[clever lintel]

Isn't GPG a protocol consisting of a suite of different algorithm, or am I remembering it wrong?

[tired wind]

to my knowledge, yeah, GPG uses a suite of algorithms. typically gpg messages / keys have magic headers though, which i don't see in any of the ciphertext-looking base64s

[minor fiber]

yea i think RSA is in it too iirc

[formal birch]

Yes, you can just take two random number as the public key but that would be meaningless if you're looking for security

[clever lintel]

Got it

[formal birch]

btw what is magic headers?

you mean the files header?

[minor fiber]

it should be that

[formal birch]

why it's called "magic headers"?

[minor fiber]

no idea

because it's magic numbers maybe ?

[clever lintel]

presumably because you can identify the file by the header, just like the term magic bytes

[minor fiber]

it's all magic

[clever lintel]

(after all magic bytes are usually a header)

[storm kraken]

I'm planning to make a full analysis of numbers frame by frame. I want to clear some things in the audio too. I don't know if it will happen on the weekend or sooner.

[warm finch]

if you want i can use izotope rx10 for audio stuff

[storm kraken]

I want to compare some parts of the letters to each other

Like is it Nine line, line line or line nine or nine nine

[warm finch]

ah gotcha

[storm kraken]

I have izotope, but I'm not that good with it. But yes, it shall be better for the job.

[warm finch]

you can use lowpass highpass filters around 250 and 5k

oh you have izotope

you can play and select frequencies with it as well

the blue highlighted thing does selecting by frequency and time

[formal birch]

RX The best

I always pre-process vocals and then import them into my DAW

[storm kraken]

Strange

I just noticed something

all the videos are 24 framer per second

Is it not 60?

nope its not 60 lol

[tiny garden]

24 isn't too uncommon tbh

[storm kraken]

number 1

what the point in making it 24 framer?

literally only Neuro on the screen

Study, only a picture

[formal birch]

maybe easier for us to investigate

[storm kraken]

I'll check now what I can.

[warm finch]

Ik 24fps is somewhat common for movies

[storm kraken]

how do I cut with it a part of the sound

[warm finch]

Wdym

[storm kraken]

complicated?

I'll just make it as a loop...

[warm finch]

like how do you hear the selected frequencies?

oh

export selection

[formal birch]

you can just give us the timestamp if you want

what are you looking for?

if you want to hear a specfic frequency band，you'll have to do a global filter and then play the segment you wanted to hear

at least i don't know there's a feature where you can do a filter on a specfic part for the track

[warm finch]

in rx you can select time + frequency so you hear the selected freq at the time

[formal birch]

i didn't know that, there's a tutorial or a instruction?

which button is it?

[warm finch]

It's this

[tired wind]

this is probably nothing, but yet another way to interpet "fill the void" in numbers II could be to find fibonacci numbers that when the individual digits are sung, they end up being 4 syllables. in which case I believe 610, 987, 2584 and 4181 are the only ones that fit

[storm kraken]

Im going crazy

[warm finch]

We all are

[storm kraken]

Listening to it in repeats now

[formal birch]

I don't think that you can really find stuff in audio

i've done that last week and still found nothing

i've cut the numbers part (lyrics) and assembled them but still found nothing(no images or hidden audio)

and no images or noise in both study and number I

[warm finch]

im assuming this isnt a clue but there's gaps around the 5k-8k range in the study video

im probably overanalyzing it

[storm kraken]

SOmeone want to hear it in repeats as well?

to confirm if it's nine or line

I'll upload files in flac

[formal birch]

where did you get the flac file?

vedal posted it or you used some software to do interpolation on the track?

[storm kraken]

Convert the sound to flac to not lose on quality

[warm finch]

pretty sure she says nine then line when i used a freq bandpass

[storm kraken]

well they all sound different to be honest

[formal birch]

you can just select the base freq and can hear clearly that it's "nine", not "line"

[storm kraken]

The first one sounds like "Nune"

I mean in 572943

Second sounds like Nine and third like line

but maybe it's my schizo idk

[formal birch]

take a rest for an hour and listen to it again

it's most likely a hallucination

[storm kraken]

https://drive.google.com/drive/folders/17j_2th7F6N12dabTCF7aOFxZhYHfeCi6

[storm kraken]

enjoy

[warm finch]

u could also have ear fatigue

[formal birch]

yes

i forget what was it called

[warm finch]

ive experienced that a lot

[storm kraken]

Listen to it and see if you hear Nine or Line

First is 9 from the "572943"

Second is the first time we hear nine\line

and the last is the second time

[warm finch]

I feel like 572|43 doesn't make as much sense because of the SoundCloud link

[storm kraken]

It's a referense

it's confirmed 9

so we can use it

but it still sound different from the other two

Now I'll check the fibs

[storm kraken]

First fib appears in frame number: 1294.
First fib last seen in frame number: 1432.
First fib, total number of framer: 138 = 5750ms

Second fib appears in frame number: 1440.
Second fib last seen in frame number: 1572.
Second fib, total number of framer: 132 = 5500ms

Third fib appears in frame number: 1581.
Third fib last seen in frame number: 1717.
Third fib, total number of framer: 136 = 5666.66ms

Fourth fib appears in frame number: 1728.
Fourth fib last seen in frame number: 1860.
Fourth fib, total number of framer: 132 = 5500ms

Do we need to add it into the document or not?

[warm finch]

you could mark it down somewhere in the number thread library

[graceful mulch]

@warm finch has leveled up! (35 ➜ 36)

[storm kraken]

Which one?

[warm finch]

https://docs.google.com/document/d/1haiT--yrMQAhPNJlqa5JU5FbEwrkOOWWBg8FHP1vXB0/edit#

[storm kraken]

I have no idea where to add it in the Library and don't want to make people mad, so if you think it's good add it where it best to be or just leave it to disappear lol

[undone maple]

"Matching all the letters 572943"
on a calculator, 572943 is EhGZLS rotated

[formal birch]

calculator?

[undone maple]

7 segment display

[shrewd trench]

Code should have been 5418008 smh

[warm finch]

lol

[foggy brook]

Hey, unrelated to the arg. But I am solving another arg and need some help with audio files. Can someone help?

[graceful mulch]

@foggy brook has leveled up! (2 ➜ 3)

[storm kraken]

?

[formal birch]

send it

[warm finch]

What do you need help with

[foggy brook]

I’m solving a mc arg thingie, this was found on a hidden button s website. It sounds like melohi but it seems off. https://cdn.discordapp.com/attachments/1004017420194828388/1116378950105567242/mellohi.mp3

[eager garden]

All I can tell you is that there is nothing in the spectrum (at least it seems)

[eager garden]

btw, u should check the file header, it looks a little weird

[tiny garden]

idk if this data is relevant at all

[clever lintel]

I suspect that discord would strip any interesting metadata

[tiny garden]

im pretty sure they only do that for images don't they?

[tired wind]

I think this is right

[thorny bolt]

572943 is odd due to its magical property that no matter how you split it you cannot make a letter out of two numbers

so meaning, the translation of 572943 to letters

will

always

be

egbidc

im just sayin i feel like that egbidc is gonna be a key to something

[storm kraken]

It can be also base64 or any other way.

[sonic pier]

Anyone tried to brute force "Numbers" AES-crypted description with a 128-bit key made up of digits only?

I don't know if it's reasonable

[hasty vale]

what does it even mean "of digits only". bits can only be digits 0 and 1.
if you mean "regular" digits, then what are you decoding them from? decimal, hex, ascii, base64?

[sonic pier]

I though decimal

like the video say : start with 2, add 9, etc...

if it's the key, it would be something like 1725 1452 6523 4875

maybe i'm wrong

[thorny bolt]

no "==" meaning not base64

[storm kraken]

what do you mean?

[thorny bolt]

a base64 is only a base64 string if it has a == or = at its end

[storm kraken]

doesn't matter

[muted goblet]

that's only if the data it's carrying doesn't have length of a multiple of 3 and needs padding

[storm kraken]

It has a system of numbers = letters

[storm kraken]

also this

[thorny bolt]

anyways

[storm kraken]

Does anybody know how tf I can make a python script use all my CPU instead of 1 core? I always get errors with this

[muted goblet]

assuming a "cracking" rate of 10e7 then checking all 16 digit numbers isn't possible, but fixing the first 2 and last 2 from the description poem could be done in about a day. I think somebody did try it, but only in ecb(?), and only checking for printable plaintext/certain magic bytes

[thorny bolt]

maybe "number 17 first" is a big clue for bruteforcing

since we can

[storm kraken]

I can try everything if I'll be able to make this python work with CPU

[thorny bolt]

prefix it with 17

and go from there

thing is it needs to be

fast

[storm kraken]

its not only 17 first

it also probably 24 last

[snow crystal]

take a look at the multiprocessing module in python

[storm kraken]

I tired it still use 1 core or gives me errors. I can try to make like 10 different python script and tell each search for 1/10 of the possible keys.

but still they allwill be on 1 core

I guess

[thorny bolt]

ok

ok

what if the length of the string we need is the length of "ABCDEFG"

[thorny bolt]

python would probably not work with this

since

its slow

you can try lua which uses only 1 thread

and with luajit you can make it pretty fast, even faster than python

(and its simple and similar to python too)

[muted goblet]

I'm pretty sure people have already posted rust bruteforcers for this

[thorny bolt]

oh that also works

rust is

as fast

as C

which is about the fastest language iirc

so like yeah

but i still think that my theory could be a promising lead possibly

[storm kraken]

I think the best thing we can do is to guess possible combinations of the key and try to find what left with BF.

[thorny bolt]

ehhhh

i feel like trying to solve it

the intended way

might work

consider this

we might be overthinking it

[storm kraken]

we have like 30 possible ways and each another 5 or 10

good luck

the only reason we didnt solved it yet is because of this

too much possible things that can be the key

do we need to use fib? yes? so all the calculations need to be with fib, ok. How we need to use the fib?

you need try all possible versions of this

[thorny bolt]

bruteforcing a AES key is insanely hard anyway

[storm kraken]

then do we need to use abcdefg

and so on

[storm kraken]

Yess

I think most people that were active just gave up lol

[thorny bolt]

is it possible that there isnt a AES key encrypting this?

possibly this is a whole

another language?

[storm kraken]

^another possibility

see how mucyh of it?

[thorny bolt]

because, it tripped off one of the base 64 decoders i used

saying

"invalid UTF-8 keys"

because

IT HAS invalid UTF-8 keys

this thing is not exactly fully UTF-8!!

meaning its either A. another language OR encrypted to the point where it spits out non-UTF-8 keys

[snow crystal]

sure. the main reason for AES is the entropy + size of the decoded base64 being multiple of 64 byte blocks

[thorny bolt]

welp that is a pretty damn convincing reason

[storm kraken]

also if you encrypt a phone number with AES it looks kinda like the base64 in the study sms library.

[dense hill]

you need to make a subprocess
with only one process you can only use one core

[storm kraken]

yea I kinda moved to other languages now

[thorny bolt]

yeah then AES is how everything's encrypted here

what does a 128 bit AES key usually look like

[storm kraken]

anything

it should be 16 characters

[thorny bolt]

what i meant

what chars is it limited to

[storm kraken]

it's just a phone number doesn't have enough so when you encode it with a 16 character key to base64 it turned to something similar like in the study one

[storm kraken]

I think anything

I guess anything on your keyboard

[snow crystal]

it can be literally any squence of 128 bit. doesn't have to be printable characters

[thorny bolt]

then its kinda impossible to decrypt it here y'see?

[storm kraken]

this is where you need to solve the puzzle

[thorny bolt]

yeah thats the trickest part

it could very well be

that the decryption key

is just numbers

or even letters

[snow crystal]

that's why AES is considered a secure encryption algorithm 😉 unless you can make assumptions that limit the possible inputs you will not be able to bruteforce it before the universe ends

[storm kraken]

I bet in the near future it'll be full of holes

[night crag]

I just hope we get some more hints. The way it is now is just to inconclusive

[thorny bolt]

new discoveries:

572943 actually matters

we now know the meaning behind "another 9" in that one lyric

we also noww know its quite literally adding another 9

[minor fiber]

Thats already been discovered from a long time you can check the docs

[azure fractal]

What's arg

[minor fiber]

Yes people did try brute force with 17 as the two first and 24 as the two last but i don't remember if they use ecb and i don't think they used any IVs or padding

Also correct me if im wrong but some website pads their keys if you enter one smaller than 16 no ?

[unique obsidian]

hey i was thinking what if the start with x end with y is to decode numbers but ive been away so excuse me if youve already solved it, also the number string could be the first time a number appears is one of the b64 numbers if that makes sense, but my code isnt working so... (im not up-to-date, so excuse me if youve kinda solved most of i)

[minor fiber]

i don't thing the number II video was that much solved so you might find some stuff for sure

[violet drift]

clue?

[clever lintel]

that the twitter account we noted earlier?

[tired wind]

at first glance this isn't @_neurosama

[ebon shale]

it isnt @_neurosama, because they arent suspended

[clever lintel]

got it, random account then

[inner minnow]

probably someone trying to sabotage, or just be creepy.

[violet drift]

yes, random account

[muted goblet]

aye, that topic was just specifically assuming a 16 numeric key. I've been trying null padding and concating the passphrase to itself till it was long enough for other wordlists myself. Bar that I could only realistically see md5

[burnt coral]

Might be a really old account that got hacked and the hacker mightve sent racial slurs or something

[hasty vale]

i am proposing to nuke "MAN#3701 take on the lyrics (instructions for block cipher)" from numbers thread library.
the sheer complexity it would take to make a symmetrically encrypted text with "and" encryption is too great to reasonably consider for even just 1 round of encryption

also i would like to kindly ask if anyone has a convenient 1040 byte string of sensible data to donate for analysing its encrypted entropy

[hasty vale]

i am pretty sure it would be possible to reverse engineer almost the entirety of the original key used for "and" encryption, and the keys for other steps from it as well, but i am also pretty sure that you would just get 128 bits of "1"s

[shell heron]

not sure how serious that question is, but:
If I need like up to 256 bytes of medium-entropy data, I usually just let my brain play LLM and type some random sentences about what I'm doing, thinking, seeing, etc.
If I need much more than that, I sometimes use something like Alice in Wonderland or other books (the projekt gutenberg website is pretty neat).

(https://gutenberg.org/cache/epub/11/pg11.txt)

[elfin swallow]

resonable, I forfeit

[rose sparrow]

You can easily reach 7.9 entropy by just compressing some data instead of encrypting it, so if you're trying to disprove that the high entropy mean encryption that's already known. Try analysing the entropy of a compressed image.

[clever lintel]

Wouldn't a compressed image still have magic bytes?

[tired wind]

I think it should, either an image header or a header indicating compression. If it really was compression, and the headers were intentionally stripped away, I think it'll be too unreasonable for a puzzle. There are so many compression schemes out there

[clever lintel]

Agreed. I'd look into raw headerless compression if we had direct indicator from the ARG that we're looking for compressed data. It's the same reason I'm skeptical about trying to decrypt the data with a specific algorithm like AES. There are just too many options.

[warm finch]

I might look into libraries/stack overflow questions related to python and cryptography since I feel like there might be something there

(to try and decrypt the data)

[clever lintel]

Sounds like a plan. I get the feeling we should be looking for a way to determine what we're even dealing with. Like how Bit recognized that the frame of garbage signs might be malbolge code

[shell heron]

okay so I wrote a program that (quite naively) generates some numbers based on the lyrics.
I didn't include the more complex interpretations of some lines yet.
It generated about 1100 number sequences, but only 75 of those were 16 digits long.
only one of that did not end up with a bad decrypt (assumed IV = nulls), but of course it's just a bunch of gibberish

[graceful mulch]

@shell heron has leveled up! (1 ➜ 2)

[shell heron]

the one non-bad decrypt was with 1789901481413904 btw

hm, entropy is 7.79[...]

I guess I gotta implement the less simple stuff

[hasty vale]

no, this isn't what i am trying.
i am trying to get average data entropy for all (some) data obfuscation methods and plug the values into markov's inequality to get the probability for data with the exact entropy as the strings we have from the method.
this would obviously mean that it would likely give more than 75% chance to data being not being messed with at all, but the point is to look for comparatively most probable approach rather than simply "likely" approach
refer to this message by Thanatos about minute differences in entropy #1112401279680008242 message

[hasty vale]

this is exactly what i am trying to do, using probability theory to cheat and find likely obfuscation method

[muted goblet]

In regard to the compression talk earlier offzip's pretty good at scrubbing for raw headerless DEFLATE streams, but nothing there. It would also be strange since all 3 ciphertexts were %16

[muted goblet]

Can you send the full 1100 wordlist?

[shell heron]

https://gist.githubusercontent.com/Dystakruul/625c3875fc69eb8c9ded046463806031/raw/a31c9521eede0eb7510dd856dac70b2b22de8757/neuroARG_Numbers_allGeneratedNumbers2.txt

I added some instruction variants, so that's the second version, which is slightly bigger (6400 numbers)

I can also upload the code if you're interested.
it's just not written with good readability in mind...

[burnt coral]

What programming language?

[thorny bolt]

did you include any fibonacci or golden ratio stuff in this sequence generator or

[shell heron]

JavaScript (Node.js, i.e. server-side JS)

[shell heron]

nope. no fibonacci shenanigans yet

https://gist.github.com/Dystakruul/0a4c82029f4faa2cf011b5c4a388b59a#file-neuroarg_numbers_generatenumbers-js

I wonder if this code needs content warnings

like
"*Beware: * r e c u r s i o n"

[ruby fiber]

just thought of this and want to throw this out there, but what if flipping the numbers backward means literally flipping them backward, and then match them with letters (ex: 3 → E, 6 → d). I've messed around with it and have gotten some results that are 16 bytes so it might work as the AES key/IV.

Example: by treating the "add the line" part as a line break and using "Add the numbers to 4" I got the number

1712478247604
58
↓
lclspcBspcdop
sB

This works as the key/IV when converted to hex, though I still don't know how the ABCDEFG part comes into play and there is still a bit of guessing as to what letter each number translates too

Probably not right but at this point I feel like everyone's a bit desperate

[hasty vale]

https://tenor.com/view/cringe-gif-24107071

dead chat.
how about a vote for keeping or letting it die. i was thinking it could get nuked by discord automatically but i guess it won't

[clever lintel]

We could just let it die. I mean, main thread isn't moving very fast now. That way we could just pick it up again if main becomes too busy

[minor fiber]

yeah thats weird it should get nuked after a week of inactivity iirc

[burnt coral]

dont lock this yet tho

[random drift]

It will get auto archived not deleted

But we can pick it back up at any time

[hasty vale]

so what are those b's

[fleet condor]

hex maybe?

[rose sparrow]

I think the numbers on the right with the distortion means that it's not the actual result, the biggest clue is what's at the top

[fleet condor]

but then we only see bs

[gritty fractal]

Possible but it’s only bs which is weird

[fleet condor]

yh

its instructions on what to do

[gritty fractal]

Maybe something like modhex? Both that, but similar

[rose sparrow]

It says that we should take all the "secret" numbers from the number 2 video, and run the algorithm from the numbers 1 video on them

[fleet condor]

yh

[gritty fractal]

Though also, where is initial_number even used?

[fleet condor]

initial_number = 2 is a parameter used for that (idk what parameters the algorithm takes so im guessing thats what its for)

for the algorithm from 1 im guessing

[gritty fractal]

The function calls are before hand

[fleet condor]

but im not too familiar with the arg atm

[rose sparrow]

the numbers 1 video itself says "start with number 2", this is just confirming that 2 is some initial value (IV) for the algorithm

[shrewd trench]

oh god what if we need to find those exact keys and note down what b is substituting

[gritty fractal]

Yeah, Im aware of that, it’s just the way it’s formatted feels off to me

[fleet condor]

yh, that could be possible

[rose sparrow]

Would be very nice of vedal to give us a way to confirm our results that way

[fleet condor]

also, what are those lines above/between letters for?

some values are also completely missing, not replaced with b

[heavy nexus]

Those b's look llike musical notes to me. God, I can't read.

[gritty fractal]

They kinda remind me of musical notation

[hasty vale]

so iv is just ...0000000000000010

[storm kraken]

we need to do a reverse I guess

[tired wind]

bs could stand for bytes in python. Like b'123' for example

[storm kraken]

secret numbers from numbers 1 to do something with numbers 2 to get the numbers we need

[rose sparrow]

Could be an IV for the numbers 1 algorithm rather than AES itself

[storm kraken]

I guess maybe the numbers on the right side are the result of applying numbers of numbers 1 to the array of numbers 2

[hasty vale]

what does that even mean

[tired wind]

Yeah how would that work

[gritty fractal]

Wouldn’t that be more likely to be secret_number?

[tired wind]

It would be like explaining an entire encryption scheme with substitution stages and whatnot through a music video

[shrewd trench]

Do we have numbers given from Numbers 1 besides the lyrics?

[storm kraken]

the instructions from numbers 1

[shrewd trench]

there's a specific "secret number" to apply and mix around with numbers 2

[rose sparrow]

It's much more likely to me that the left column shows the parameters and the right shows the heavily distorted output

[minor fiber]

from a programation stand point applying cipher.number1(secret) to cipher.number2.array() doesn't really make sense to me

what is cypher here ? a superclass ? does number2.array() return an array of object which have the method number1(secret) ? do we have an interface here somewhere ?

these pairs of numbers and stuff id strange

since generating keys is called only once i guess it's call by cipher, then we call on each object the numbers1(secret)
they return a pair of number with results, some contains letters some don't, we never got letters from our attempts with number1
i guess we do see a lot og 24 at the end so it clarify the add the numbers 2 4 (lmao to 4 believers)

tbh the "b" might be corrupted data just like the lines or weird characters

[hasty vale]

wait, so numbers 1 is indeed an encyption algorithm instructions nooooooooooooooooooooooooooooooo
i dont want to implement and encryptions

[storm kraken]

maybe the key from numbers 1 include numbers and the letters "abcdefg"

this is why in generated keys we see "b"

[rose sparrow]

cipher is an implementation of the algorithm that's hidden from us

[gritty fractal]

Yeah, yeah. But for a initial number for numbers one algorithm wouldn’t it be secret_number as it’s put as a parameter in the pseudocode?

[hasty vale]

it's like that in pycryptodrome, you create cipher with an algorithm, then you call that cipher's methods

[gritty fractal]

Yeah, and as I’ve said from programming standpoint defining initial_number after encrypting doesn’t make much sense

[minor fiber]

from what i can see the distortion can be either :

• "b"
• number erased
• weird artefacts

so we can gather that this algorithm do return in a range of 17...24

[rose sparrow]

You're seeing the printf output, not the psuedocode, the order of the lines don't reflect the order of operation

[gritty fractal]

Oh fair brain go brrr

Study js confused me

[rose sparrow]

It's much more likely the cipher is being run during the "Generating keys..." line

[tired wind]

If we're running by Numbers I being instructions for an encryption algorithm, initial number should be the inital state. Then the keys we see below are probably what we should aim to get?

All speculation, probably wrong

[minor fiber]

problem is keyword "keys"

[hasty vale]

maybe we need to go through numbers 2 description in 128 bit parts applying it to cipher one by one

[minor fiber]

personnaly i tried doing the algorithm from number 1 starting with the numbers of days as starting number (might not have done the algorithm correctly) but i got no result with it

[minor fiber]

i speculate it's because of distortion

[tired wind]

This sounds like a good idea

[minor fiber]

if you look at the distortions i listed every result are in range of 17...24

[rose sparrow]

Ah yeah you're right

distortion threw me off

[robust scaffold]

did not see anyone saying this and I'm probably wrong:
maybe "public static void" generating things are study-sama? attempt to solve something (the descs?) with number1 algoritm
so we can check if we have the good algorithm from number 1 with number2 numbers and try to find the starting number somehow

[sacred ruin]

okay

i have big theories

the symbols may have their meanings

for the weird music note thingy, you can see it appears mostly at the start

[minor fiber]

yes

[sacred ruin]

like x7

etc

that can be a 1

then theres a little _ but at the top

mostly at the end

which can be a 4

the keys that have that, are the ones we should focus on first

[minor fiber]

there is also _ at the start which replace a 7 tho

also artefacts seems to replace sevral numbers

[sacred ruin]

4913 may be one that we need

either as IV or key

then 4156

either as iv or key

[hasty vale]

what about when the line is above a number instead of replacing it

[minor fiber]

the fact that most results are 17590...24 and 17589....24 is intersting tho

[sacred ruin]

we should do those first

our main priority is finding out how to turn those into 16 bit keys

the "b" could be 11

hex

which would invalidate 4156

[minor fiber]

i don't think this is substitution like that

there is so many numbers being replaced by each one

[sacred ruin]

4913: 175b068887b24
4156: b7189938327924

[minor fiber]

i'd rather we focus on the numbers that are readable on one group

17 590 24 are repeating int he first 2 group

even third

[sacred ruin]

even if we replace the b with 11, 4156 gives us 15 numbers

and so does 4913

[shell heron]

I ain't reading all this but I'm sure that "add another line" means add another "1"

[sacred ruin]

Add another 9 yeah
Add another line yeah

we can add "1" after a 9

[minor fiber]

the 1 would go through multiplication and reversing

[shell heron]

I did the thing backwards with one of the examples
I ended up with the correct number

the b's are placeholders, just like the musical symbols

[rose sparrow]

"correct" number?

[sacred ruin]

weirdly enough it always ends up with a length of 15

[shell heron]

I'll try to type it out

[minor fiber]

because they each have at least one distortion so can't be 16

[sacred ruin]

175110688871124
117189938327924

[minor fiber]

distortion can change more than one number

[thorny bolt]

oh hey you guys moved here, time to lurk

[storm kraken]

It says 1 is 256 bit

[minor fiber]

binary

[rose sparrow]

So what do we do with this?

[storm kraken]

Nothing it's desc from num 2

[shrewd trench]

capital letter is replaced with ?

[rose sparrow]

Is 26 a reference to "matching all the letters"?

[storm kraken]

Or

[shrewd trench]

doesn't look anything like Y

[rose sparrow]

The Y is in front of the Z

[shrewd trench]

Oh wait i see it ignore-

[storm kraken]

It can means that x/26Yz is not part of the code.

[shrewd trench]

It's jusut really hard to see "Y" in front of the weird blurb hovering in the same area

[storm kraken]

In study it says xyz

[deep flax]

I thought the Y is 8

[tired wind]

assuming we get a key/secret number, we still have to know the following to verify:

1. What is the cipher?
2. How exactly to "apply" one to the other?

So we are technically working with 3 ciphers/derivations here, the one to get the key, the one to get ciphers.numbers1 & ciphers.numbers2, and the one to apply both of them together

[shrewd trench]

It's straight up the description from numbers with strange artifacts that mean nothing

[sacred ruin]

the cipher is numbers 2

the decryption of numbers 2 is the key for numbers 1

[storm kraken]

What the lyrics of numbers 2 again

[sacred ruin]

then from there we do the soundcloud and studies

[storm kraken]

Take between x and y and odd one out?

[thorny robin]

Numbers that we can see in the new video

[storm kraken]

No I mean the xy thing

Start from X. End with Y. Remove XY + odd one out

Odd one can be z

So we get /26

Decide to 26?

Devide*

[tired wind]

Probably, the issue is with what exactly those encryption/key derivation schemes are. Are we expecting the same scheme for all decryption/encryption or will they be different? If different, how?

Are we postulating that the numbers videos themselves are the respective schemes?

[thorny bolt]

x/268z

x divided by 268 Zs?

[storm kraken]

It's not 8

It's Y

[thorny bolt]

oh damn it

[rose sparrow]

Hmmm, divide the numbers from Numbers2 by 26 before inputting it into Numbers1?

[sacred ruin]

no

wtf

you clearly dont know the arg

[shell heron]

4156  ?7?8993832792?

?7?8993832792?
?8993832792? ("seventeen is first": remove 17 from the front)
?899383279 ("add two, four": remove 2 and 4 from end)
?899383279 ("multiply by 9": divide by 9. this is where the first number is important. there will be two possible values that divide evenly and one of those will be ruled out later. for now, just assume 5 as the first number.)
-> 5899383279 / 9 = 655487031
655487031
655487021 ("make the two a three": replace 3 with 2)
120784556 ("flip the numbers backwards": reverse the order of the digits)
12078455 ("add another six": remove the 6. this also lets us know that "another" doesn't require that digit to be already present in the number)
2415691 ("multiply by five": divide by 5)
now the next instruction is "add another nine, add another line" and the numbers at the end are 9 and 1, so "line" must refer to 1
24156 (removed 9 and 1)
4156 ("start with two": remove the 2 from the start)

AND THIS LAST NUMBER MATCHES THE FIRST NUMBER

[shell heron]

fuck yea, it worked a second time

[storm kraken]

If we take from x to y it's /26 and we left with z, which is probably the odd one. So we left with /26 if we assume that x/26Yz is part of it.

[sacred ruin]

that does make sense

[sacred ruin]

mate

didnt you just see it

its the start of the cipher

its not a clue

[storm kraken]

I know

But you have the instructions and it appears in the audio

[sacred ruin]

thing is, can we get the full number this way?

[hasty vale]

so this is numbers1 cipher

[sacred ruin]

17589938327924

if we take your logic this is the number

ow

[hasty vale]

ok, we have a numbers 1 cipher, what now
what else to plug into it

[shell heron]

the solution to Numbers should be 1759032572147924

[graceful mulch]

You have unlocked new role

[shell heron]

if i didn't make a mistake

[sacred ruin]

okay

cook more

im liking it

we need a second key now

@shell heron can you do 4913

[robust scaffold]

just asking, do it work with others numbers like 201 or is 4156 the only one you tried

[shell heron]

idk, haven't tried it out yet
I'd code it. I don't like doing things like this manually

[sacred ruin]

please

its urgent

[hasty vale]

nothing for any strings

[minor fiber]

problem is most don't give a 2 after multiply by 5

201
2201
220191
1100955

[rose sparrow]

How do you get this? 5899383279 / 5 = 655487031 wouldn't you get 5899383279 / 5 = 1179876655.8? Maybe you made a typo?

[tired wind]

Based on the 128bit clue I suspect we won't get plaintext just yet

[graceful mulch]

@tired wind has leveled up! (9 ➜ 10)

[hasty vale]

a = '2' + a
a = a + '91'
a = str(int(a)*5)
a = a[::-1]
a = a.replace('2', '3')
a = str(int(a) * 9)
a = a + '24'
a = '17' + a
print(a)```
code? fact check plz

[minor fiber]

yes

[rose sparrow]

Found it, the typo is / 5, should be / 9

[minor fiber]

are you running it with the list we got ?

[hasty vale]

201 - 175031009924
1321 - 1750315940924
does this look correct?

[minor fiber]

no

there should be a 9 after 5

unless 9 is a distortion we didn't expect

[small meadow]

forgot to add the 6

[minor fiber]

oh god

even i forgot how to read yea the 6

[shell heron]

yea sorry about that typo

was writing too fast

[minor fiber]

a +"6" before flip

[hasty vale]

1321 - 17590315940924
does this look more correct

[small meadow]

another small typo: replace 2 with 2

[minor fiber]

yes

[sacred ruin]

okay

[sacred ruin]

we need 2 keys

[minor fiber]

pfcpmf can you try 130 , 129 , 1432 please

[robust scaffold]

i done python for it and it work for 4156, 906, and probably 716

[shrewd trench]

[tired wind]

Do they all produce 16 charas

[sacred ruin]

once we have 2 16bit keys, get the numbers 2 description and pass it thru AES CBC, and try both keys for iv and key

[rose sparrow]

a = '201'
a = '2' + a
a = a + '9'
a = str(int(a)*5)
a = a + '6'
a = a[::-1]
a = a.replace('2', '3')
a = str(int(a) * 9)
a = a + '24'
a = '17' + a
print(a)

This gets much closer to the distorted results and is a simpler interpretation of the lyrics

[sacred ruin]

it should give us a 32 bit key which is then used to decode the numbers 1 description with AES ECB

thing is

numbers 1 has [BINARY] in it

so we may be looking for a bin file

[tired wind]

Yes

[sacred ruin]

or a binary number

[hasty vale]

the list

 (1321, '17590315940924')
 (5831, '17590356737924')
 (723, '1759032467924')
 (743, '1759032557924')
 (4913, '17590368887924')
 (879, '1759039406924')
 (875, '1759037606924')
 (2145, '17590343309924')
 (716, '1758993677924')
 (906, '1758989186924')
 (4156, '17589938327924')```

[sacred ruin]

are all of those 16 bits

filter it out

only 16 bit

[tired wind]

Seems to either be 16 or 18

[sacred ruin]

wait

[tired wind]

Seems distinctly to be 2 sets

[shell heron]

201      1759031009924
1321     17590315940924
5831     17590356737924
723      1759032467924
743      1759032557924
4913     17590368887924
879      1759039406924
875      1759037606924
2145     17590343309924
716      1758993677924
906      1758989186924
4156     17589938327924

I think it checks out

[sacred ruin]

numbers 2 had this weird binary thing

with the neuros

[sacred ruin]

can you say which ones are 16 bits

[hasty vale]

wait am i stupid, none are 16 symbols

[tired wind]

Combine the fronts

[minor fiber]

572943
257294391
12864719556
65591746831
1759032572147924

[tired wind]

Prepend 201 to 175..., 1321 to 175... etc

[sacred ruin]

at the start?

also another pattern

it starts always with 175

[hasty vale]

('723', '7231759032467924')
('743', '7431759032557924')
('879', '8791759039406924')
('875', '8751759037606924')
('716', '7161758993677924')
('906', '9061758989186924')```
only 16 symbols

[minor fiber]

yes that was said earlier

[sacred ruin]

sometimes 1759 and 1758

[sacred ruin]

okay

[shell heron]

none of the numbers from the new video are 16 bits, but if you do the same with the numbers from the Numbers video (5-7-2-9-4-3), you get one that is 16 bits

I posted it earlier

[sacred ruin]

how many possibilites are there to calculate

all of them are both tested for iv and key

wair

try the last 2

they're the odd ones

aes, 16 bits, cbc using numbers 2 desc cipher

use both as iv and key

[tired wind]

Pfcpmf and your almighty tool plz

[shell heron]

I'm certain that the numbers used in today's video are only there to show us how the 'cipher' works

[minor fiber]

why are we focusing on the examples when we figured out how to used the numbers 1 algorythm ?

thanks god someone that speak sense

[tired wind]

I agree with this

[sacred ruin]

exactly

we need to apply it to the numbers 2 numbers

[minor fiber]

no

[rose sparrow]

that's what we already did

[minor fiber]

we need to apply it to number 1

[sacred ruin]

572943?

[minor fiber]

we already did that

it gives 1759032572147924

[sacred ruin]

16 bits?

[minor fiber]

bro

[hasty vale]

nothing for any of the those numbers, but that's with zeroed iv

[rose sparrow]

I mean the screenshot explicitly says they're "keys"

[random drift]

👨‍🍳

[sacred ruin]

we literally need to use them to decode

[minor fiber]

we already used them to figure out the algorithm

now we're back to numbers 1 but we actually know what we are doing

[sacred ruin]

congrats we figured out the algorithm

[robust scaffold]

crazy idea:
we know the cipher but we need a starting number to supposively find a key
maybe use the cipher on Fibonacci Sequence until we get a 256 bits solution ?

[sacred ruin]

but what do we use the numbers for?

oh wait!

[stray crater]

this has 16 characters though

[sacred ruin]

to decode the aes!

whats taking so long testing the numbers then?

you applying algorithms to neuro's forehead?

[minor fiber]

we need a starting numbers that is given by the numbers 1 video

[sacred ruin]

no

[minor fiber]

oh god

[sacred ruin]

we start with numbers 2

[rose sparrow]

AES has very specific key lengths, need to find a way to construct either 128-bit or 256-bit keys out of the results

[minor fiber]

🤦

[shell heron]

if anyone wants to implement it themselves:

//[number] is the input, [n] is the output
n = "2" + number;
n += "9";
n += "1";
n *= 5;
n += "6";
n = n.split('').reverse().join('');
n = n.replaceAll('2', '3');
n *= 9;
n += "2";
n += "4";
n = "17" + n;

this is probably close enough to pseudocode

[sacred ruin]

you probably havent seen the proof we got

numbers 1 is a 256 bit aes cipher

[minor fiber]

pls darkeew just breath a sec

[sacred ruin]

where can we get a 32 bit key

[tired wind]

Good idea, but I believe the 128-bit clue suggests we should get two 128 bit blocks. I could be misinterpreting tho

[sacred ruin]

decoding the numbers 2

1759032572147924

[rose sparrow]

We did it in python before

a = '201'
a = '2' + a
a = a + '9'
a = str(int(a)*5)
a = a + '6'
a = a[::-1]
a = a.replace('2', '3')
a = str(int(a) * 9)
a = a + '24'
a = '17' + a
print(a)

[sacred ruin]

can someone test that key to shut me up?

prove me wrong

[minor fiber]

we have an algorythm that take a number and return a key, so in the numbers 1 video we need to figure out the number this algorithm takes

[sacred ruin]

okay then

prove me wrong

[rose sparrow]

Do it yourself

no one has to prove anything to you

[minor fiber]

there is no wrong we just don't have the number to give the algorythm

[sacred ruin]

i would but im on mobile

im asking for help

[hasty vale]

only 10^6 possibilities to bruteforce

[minor fiber]

no

[sacred ruin]

thats why we have a thread

to help eachother

[rose sparrow]

works on mobile https://gchq.github.io/CyberChef/

[hasty vale]

test where

[sacred ruin]

aes decryption

use a tool online

[minor fiber]

we need the first line a = [number]

this unknow number need to be found somewhere on the first video

[hasty vale]

i already did. or you mean with ivs?
already did that

[sacred ruin]

numbers 2 desc and 1759032572147924

bleh

[minor fiber]

the whole point of these additional hint were to help us with the first video, we didn't know how to use the lyrics now we do, but we need a number to use the lyrics on

[sacred ruin]

wait

the thing

the photo from the newest video

check the first line

[minor fiber]

why look at the newest videos

[rose sparrow]

I think we're using it on the right numbers, we need to construct keys from them

[minor fiber]

i don't think so we using it on examples we were given

[stray crater]

guys

[minor fiber]

we have to remenber at the start we only had the numbers 1 video so the ARG can be solved only with the first video

[shell heron]

using 1759032572147924 as a key for both video descriptions results in bad decrypts
* with IV of all zeroes

[sacred ruin]

no

[minor fiber]

if we keep forgetting the first video and focusing on the new ones we just gonna keep being stuck

[rose sparrow]

That's not true, tons of ARGs that can't be solved with just the first clue. In fact it's often done intentionally so that the ARG isn't solved too quickly

[stray crater]

I did Key and IV as "1759032572147924" > CFB > Input Raw > Output > HEX and with the input being the Numbers 2 gave me a wall of numbers and letters

[minor fiber]

okay but do you agree that all the new video are clues to help us understand the first video

[sacred ruin]

?

[stray crater]

no screenshot because embed permissions

[rose sparrow]

To help us understand the first two I'd say yeah

[snow crystal]

I (think?) I agree with Amarok. We got the cipher.numbers1() part of cipher.numbers1(secret_number). Now we need to find what that secret number is...

[minor fiber]

i mean yea study is just a whole bag of stuff too i agree

[gloomy cedar]

given all the golden ratio references could the secret number be 1.618

[stray crater]

Can anyone confirm if this output means something?

[minor fiber]

we didn't explore the golden ratio that much so

[rose sparrow]

secret_number is just a parameter for a function, vedal is a programmer, the first line in the screenshot clearly says that the secret numbers are in the numbers2 video/array

[thorny robin]

Send screenshot to my dm

[minor fiber]

also from the lyrics it seem this secret number as a 9 and a 1 and when multiply by 5 it has a 2.

"add another 9 add another line"

[rose sparrow]

Your output is hex, the wall of text and numbers is just garbage represented as hexidecimal numbers

[shell heron]

I'm gonna go to bed; it's way too late here
good luck guys o7

[stray crater]

oh, ok

[sacred ruin]

this video is literally all the clues we need

spectogram said numbers 2 was 128bit and numbers one is 256 bits

well

i guarantee this wont be the end of the arg even if we solve it

pretty sure the soundcloud isnt just for giving clues..

[rose sparrow]

stop wildly speculating Darkeew, this channel isn't about that

[sacred ruin]

and the study video number

[rose sparrow]

need to focus the conversation in this thread on the task at hand

[sacred ruin]

do it then

im not stopping you

[rose sparrow]

but yeah I agree that "1st = 256-bit [BINARY] 2nd = 128-bit" is a clue to constructing the key

[sacred ruin]

im speculating here because most of you dont actually check the topic at hand in the other thread