#Neuro ARG
1 messages · Page 40 of 1
Also when you decrypt you might want to just check for valid UTF so you don't end up with 16GB of data
And if there is no valid UTF in 10^12 rang then idk heck
smh my head, people missing my api results
why not calculate entropy on the go, same with binwalk
I did use Binwalk
It was actually binwalk I was multithreading https://gitlab.legitcorp.com/neuro-arg/local-devglan-decryption-script/-/blob/generic-decryptor/main.py?ref_type=heads#L54
yeah actually the results I mentioend don't apply to AES decrypting, sorry
I believe it's numbers, so I check for ascii.
That was so quick I didn't bother, it's the entropy and file signature analysis that was slow 
you do, but with handles it for you
I couldn't get binwalk to work when passing through a string
Don't know binwalk so can't help on that
Technically you can also pass binwalk a list of files instead of doing it invidiaully but that didnt' work either, I think Binwalk might just be buggy and unmaintained 
Probably can get better results with something else if you know one

I don't think it would be too difficult to just lift binwalk's entropy analysis out of it's source and use it as a local function against decrypted
Not sure exactly what goes into it, but it feels like the kind of calculating that could be offloaded into a C library
If anyone knows C and python and python libraries cough cough 👀
Thanatos always pulls through 
Entropy is just a calculation of how random something is, random data tends to be encrypted or compressed, idk how it's calculated normally
maybe i am just stupid and dont understand english but isn't it weird that aes decryption even for just cbc is faster than entropy calculation, especially with a dedicated library
i used my own "modified" version of thanatos' entorpy calculation and aes encryption which is supposed to be faster than decryption and aes still took like 80% of runtime compared to 10% entropy
now I'm confused. you said decryption is faster than entropy and then you said aes takes 80% of runtime while entropy takes 10% 
I didn't say that, baa said that
That was so quick I didn't bother, it's the entropy and file signature analysis that was slow :nooo:
which i dont get.
surely it would be more efficient to get binwalk and entropy calculation along with aes decrypt rather than get 16gb of decrypted data
Are you guys still trying to brute force aes?
ah ok. didn't realize you were referring to that
@dapper wraith has leveled up! (5 ➜ 6)
also i guess since its a tiniest bit relevant, i'll post my entropy calculator. it's not amazingly faster, but it basically rides on precalculated math.log
from collections import Counter
def entropy_real(data: bytes) -> float:
global precalculated_dict
entr = 0
data_len = len(data)
counter = Counter(data)
try:
for x in counter.values():
entr += precalculated_dict[data_len][x]
except: # fails if length isn't in precalculated_dict
for by in range(256):
p_x = counter.get(by, 0) / data_len
if p_x > 0:
entr += - p_x*math.log2(p_x)
return entr```
and json with precalculated entropy for lengths 256, 400, 1040 and 1264, if anyone wants to use it ||(no one does)||
you could probably save a lot of time if you ignored decrypted data that contains non-printable characters
two integer comparisons per byte and you won't have to check the remaining bytes most of the time anyways
I don't know, maybe the AES library is written in C so much faster than my entropy calculation, or it's just overhead from reading and writing, or it might be the signature analysis that takes a while,
But what if it decrypts into a ZIP file 
That would just be cruel 
yea
but you could also just check for a handful of common magic bytes
it's faster than checking the entropy of 99% garbage
yeah not a bad idea
and if there are multiple levels (of encryption), we have no chance anyways if all we do is bruteforcing
i am literally using the same cryptodome library as you do
tbh the language doesn't matter if your method is slow
I guess you can compensate for both by throwing money at it though
entropy seems pretty negligible in my setup https://gyazo.com/5593f2d8ccf5663a6e13c6eeffcce0b3
also while this (along with try: .decode() except: ...) is sensibly true, no one says the entire message must be valid.
it very much could be the case the text is padded with garbage bytes, and this method would not catch it, especially if you're only doing entropy calculation on magic numbers hit
i was thinking, since the initial data is so much more impactful on the resulting entropy of the encrypted string than the encryption method, maybe i can start see scores for different methods cycling through strings.
basically, look at the data and the results will surprise you. removing only keys, aes ecb absolutely dominates, the scores of other aes methods don't even compare
I don't think I understand what you're doing or what you're trying to show, but I keep thinking about the fact (?) that AES-ECB's ciphertext will 'inherit' patterns of the plaintext.
which probably makes the entropy correlate more strongly? idk
is the thing you're saying "[1] the Numbers-ciphertext has lower entropy than average AES-CBC ciphertexts (plus most other modes) and [2] AES-ECB is more likely to have lower entropy compared to other modes, so [3] the ciphertext is probably a result of AES-ECB encryption"?
Okay, managed to improve the code by 10-15%. Should run faster, maybe I'll manage to do 10^12 in 24 hours instead 30.
what i was trying to say, is that if we assume all ciphertexts are results of the same encryption method, and since data showed much stronger correlation between (initial-text to entropy-of-the-encrypted-text) than (encryption-method to entropy-of-the-encrypted-text), it would make sense to cycle through different initial strings and matching them to numbers 1 and numbers 2 (as opposed to just previously assuming both texts have the same contents, like both hold a jpg file or both are latin text)
so with these initial assumptions (along with the assumption that both texts aren't literally just random bytes), calculating the scores based on probability of each entropy value occuring, aes-ecb method gains massively higher scores compared to other methods, so [3]
different plaintext content of the same size has effectively no bearing on entropy of ciphertext from modern crypto. Modern crypto algorithms should also have the same entropy compared to each other with the exception of chaining modes, where repeated block size plaintext in ECB would have lower entropy. That would only show up if there were any repeated blocks in our ciphertexts though, which there aren't
yes, as thanatos said.
the entire idea was dead on arrival, it really failed to fulfill its initial goal which was to filter out unlikely encryption methods, which as can be seen, it does not do.
one could even say it made negative progress.
which is to say, not to offend anyone by implying they ever did, don't listen to me
The decryption efforts (including brute forcing and identification) are definitely valiant, but at this point we are really only doing it because there seems to be no other workable leads
I'd say it was a lost cause the moment we started bruteforcing
surely the new video will give us all the information we need
surely

It's nice looking back here after a long time of forgetting about the ARG and seeing all the progress we made

i have
a
theory
recently java removed the need for public static void
this MUST be referencing it

removed the need for public static void ?
i heard too that java 21 is "moving towards" simplification by removing the need to explicitly specify things, i guess public, static or void being one of those things.
this video giving any answers is massive copium, because vedal leaves for a week right after, which probably means he doesn't expect anything to happen within the week that he is gone, it is just going to be numbers 2 again
@austere sun has leveled up! (10 ➜ 11)
so what we gonna have then ? just main()
how do we declare functions anymore, tbf people don't use java newest version anyway they usually stay with the old one
They had the chance to make a top level main file like C#, but instead they chose to have void main(){}. Now everybody's gonna hate it.
ok here is something promising
the fact that this video
has no description
there is no description on the video
no base64
if this is the case the entire video might be the only thing
how is it promising, this is horrible, we are (i am) starving for info.
also yeah, the api call did also say that there is no desc
maybe having less info in 1 video is more helpful, kind of like getting us back on the right track, perhaps?
since people could easily be overloaded by
all the new info
void main() {
System.out.println("Hello, World!");
}
this is how you make main functions now i guess (in java)
like we get a new base64 (possibly AES) string, then another guide music video thing
we're obviously gonna think its for said base64, AES string in the description
but what if it wasnt? there isnt anything proving it is
the video could very well just be a thing to get us back on the right track to solving things, maybe?
im wondering is there a way to get information about the algorithm if we have sevral of these Base64 ?
idk if there is magic bytes that we can compare to see which aes it is or something

so more Base64 isn't helpful then
so much new info i am practically drowning in content 
next 2 weeks are going to be literally this again
what else is it then, pick a normal distribution graph, as you can see they are so distinct and informative

custom Base64 is just a straight line 
I don't think we can say for sure that the description would be empty by the end of the premiere though.
maybe it'll be updated manually, or a script has been rigged to do so. The puppet master does know programming after all
It's possible if each video description was an OneTime-Pad ciphertext that had all been encrypted with the same key text. If you XOR the ciphertexts together then because the key has been applied to both, it will essentially be XORed away with itself, leaving you with the XOR of two pieces of cleartext. With some statistical crypto analysis I have long forgotten and probably never fully understood, you'd be able deduce what the key is and get two cleartext detangled. For each additional piece of ciphertext OTP'ed with the same key it'd be easier and easier to figure out what the key is.
Based on that, it might be possible to deduce if it's OTP with the same key (which isn't actually true OTP because it's a mistake) if we XOR the ciphertexts and measure the entropy.
But that's literally the only algorithm detection I think would be possible.
Modern ciphers are designed to not give away info.
Though of course they might also leak data if they weren't implemented correct eiher.
good to know
has anyone checked the sc?
checked for what?
nothing new there
Any breakthroughs with the third video?
CLUE schitzo
there is three video for a while yes "Numbers" "Study" and "Numbers II"
there is a fourth video coming up
is it possible that the voice in the videos is v3? :schizo:
world.execute(me); is playing
might be some clue :schizo:
guys doesn't this sound a bit different than last time 
because v3 
don't make me go all schizo on you again, david
who are you talking to bread ? 
myself 
understandable
huh so they have stronger CPUs... 2 of my borrowed pcs done with the keys in like 16 hours instead of 24/30. The other pc still doing the job...
But these pcs also cost double the money.
what if we just made a way to share all our computer resources into this thing
no point in this, people bruteforcing on their own is as good as it gets.
the thing is, even with 100 computers bruteforcing, the sheer amount of possible combinations is too great
I'll share all my files when I'm done, maybe it will help but I'm doubt it.
If the whole world will combine all the CPUs, maybe we will be able to brute force it fast enough to not die before.
simple really, get a quantum computer

The only reason I managed to do it, is that I'm considering 17 first 24 last and I have only 12 numbers between.
The current quantum pc are kind of not really a quantum pc.
imagine if like
And I doubt it will be enough
none of these hashes are suposed to be AES or Base64
but some custom yet easy encryption that is gonna be explained in the new video
really thinking we can solve it with new video within 10 hours before vedal leaves for a week 
I thought about it from the first day, but the best guess we have is aes or something similar.
If it's indeed something simple, we need to figure out the shortest code of study.
Which is highly likely to be a mobile phone number.
I think it's possible
instead of reacting you could respond ❤️ ❤️
if i respond to everything at that rate im going to have more messages on the server than the server owner and that would be pretty cringe.
also numbers thread got hidden. it's over. billions must bruteforce
Numbers thread isn't hidden for me yet?
the most confusing part of this arg is who's singing
yes
"no" ok who is it then
😁
cuz it does not sound like neuro
people thinks all these videos are the only thing to solve when truly it's just the tutelrial
I'm here for the slander sequel video
also
the subtitles may not be the actual subtitles
they can be uh
an entire base64 string...
fcuk, true, they really can just be more encrypted strings rather than actual subtitles

going to keep api call ready to pull them immediately then
am I schizo or it looks like something neuro can say with the exact emotion...
die >c<
thats the "im an engineer" face
lore confirmed, she lost her battle
R.I.P
lol
it's probably just a coincidence, but still a funny and strange one.
Hmm some keys with the right IV maybe can give a phone number. I just found one and played a little with the IV and managed to get some numbers with letters, but it was like almost all letters before. From this: "+44_pc<?0Xsg" to this "+44Vt`973Vtg"
If the IV is different than the key, it'll be much harder...
are we going to have a group watch of private static void, because if we do, and if i am going to record it too, i will need to set up separate audio channels in obs which seems to be as painful to do as it was 5 years ago, so i would like to know in advance the plan
afaik you can use youtube-dl to download premieres as soon as they start
(not sure about the state of ytdl tho. maybe one of the big forks would be better)
wait so that static neuro discord account WASNT a fake?
I think we should
The title of the upcoming video has nothing to do with static sama
Also static sama was a name we made up for the fake acc
Folks, I've made a script to download a YouTube video's subtitles, description, video, thumbnails, extract the frames, and also brighten them.
https://gist.github.com/jameshi16/16c2569a92eaa91e8512ad096195d78e
Hopefully this'll save us some time when the video drops
Oh, nice. I did it manually.
where's the script that solves the arg
Keep dreaming
neurosame
wow isn't that colossally convenient, that all of that can be done by basically passing a checklist to 2 libraries and they just give it to you, not even any authentication required
i wont bother with recording then
ok, maybe it isn't as convenient as i thought it was.
you can't make pull requests on gist, can you? not that i would know how to fix this without just removing multiprocessing
that's pickle
that's object object
that's actually pickled
that's actually object
that's pickle
that's object object
that's actually pickled
that's actually object
Crunchy pickles
ok i made it not throw lambda error, but now it's just empty 
i am going to wait for vanor to fix it, because i have no idea how any of this works
Btw, there is a way to skip most of the timer at the start of the premiere
Just either press L key or set speed to >1x
I've heard of salting, peppering and currying, but what the fuck is pickling??
oh it's a python thing
you usually pickles stuff like vegetable to increase the shelf time
it was vastly used before fridge were invented
can I pickle neuro plushie?
apparently the video needs to have third party caption contributions enabled to download captions from api, so that sucks.
also the way youtube-dl formats captions sucks, especially if the captions are going to be base64.
at this point might be easier to just wait for premiere to end and just copy paste the transcript, hoping the video isn't privated immediately after, surely it won't be that cruel 
the way youtube-dl formats captions sucks
it's just a very standard subtitle format
you can play a video in VLC, add/import the subtitle file and you have working subtitles. just like that.
it's annoying if all you want is a transcript, but those file formats are designed to be used with a video to... you know... display subtitles
:^) What version of python is this
3.11.0
look at comments under, there are some other things
7 hours to go
Thanks. I'm running ubuntu with python 3.10. In fact, I developed the script with numbers 2 as the testing video. Certified "works on my machine" moment
The issue with the missing '6' is very weird. My guess is how Windows deals with paths, using \ instead of / causing something to happen
should work now, give it a shot. might want to delete the out directory
regarding the transcript/subtitle thing, there's a snippet someone shared to strip away all the metadata. https://superuser.com/a/1541680
we should be good on that front
yep, works now.
no compression artifacts, nowaying. this youtube-dl is insane, how hasn't it been taken down by youtube yet, especially since it does things youtube asks for oauth2 for free
embed failure, cant even bmp
for the sake of the docs, don't upload all 1500 * 3 frames
just one thing, the gamma frames have pretty bad contrast tbh.
my opencv attempt isn't much clearer, but i think it would be better if it didn't increase brightness as much
youtube-dl is insane, how hasn't it been taken down by youtube yet
not youtube but it got DMCAd by some music label a year or so ago because you can download music from youtube with it
Yeah I agree. The aim here isn't to produce something nice to display, though. Just knowing something is there and fairly readable should be enough for someone to get the original frame and make better adjustments
I'll probably play with the settings for the next lore drop (if any), or even integrate with your opencv script if you don't mind sharing
Inb4 we make a full blown arg solver pipeline
Our reward for solving the arg is the green checkmark on our pipeline
maybe vedal will star your git repo if you solve it 
it's just this. but it takes opencv image, idk if it can integrate with ffmpeg output
def increase_contrast(img):
alpha = 6.0 # Contrast control (1.0-3.0)
beta = 20 # Brightness control (0-100)
adjusted = cv2.convertScaleAbs(img, alpha=alpha, beta=beta)
return adjusted
image = cv2.imread("1560.bmp")
cv2.imwrite("1560_1.png", increase_contrast(image))```
thats the
why ffmpeg
why youtube-dl
it isnt hard to download the raw data of the video
something better than youtube dl even
i am pretty sure some of the things youtube-dl does are illegal to be that easy, how can it be easier than that
just put the ffmpeg output as np array and pass it thru opencv
its easy
you mean open the file after ffmpeg creates it and pass it through?
wouldn't that be massive overhead
what image it creates
why .run()
take that out
just make it a variable
image = ffmpeg lbs l;fksd fsdf asdfa
image = np.array(image)
easy
but ffmpeg lbs l;fksd fsdf asdfa is not an image, is it?
from what i understand the run() is there for ffmpeg to go through the video that's given to it as a parameter, you can't just 'catch' an image in between that like image = , can you
then
where do the images go
does it download them
surely you can save the output in bytesio right?
into out_file as .bmp? it's all within ffmpeg
then theres nothing you can literally do
if you cant control what ffmpeg does with the images
not sure if it helps but they got some examples on how to read video and video frames into buffer/numpy
https://github.com/kkroening/ffmpeg-python/blob/master/examples/ffmpeg-numpy.ipynb
Printscreen + Photoshop 
Banger soon <t:1687474800:R>
I tried to make text as visible as possible, in the future I may not make it as bright as it is now in the file on the disc
If you want
you rewrote the code? or what do you mean.
that ffmpeg to numpy array seems so complicated, i'd rather just take the overhead from writing and reading files, or just use opencv to extract frames in the first place.
btw i am still waiting for that better youtube-dl alternative
extract frames and brighten them automatically.
btw, am i being massively scammed, why isn't numbers 2 desc a multiple of 16, can anyone confirm
I just don't understand why we need thousands when we can highlight one general?

It is of course good to search for them with code, as they may go unnoticed, but we don't need all of them to show people in files
Or am I missing something?
it's just for looking, without uploading to docs.
i don't remember where i copied my string from, but it was either from pins or from docs, so that means one of them is an impostor 
Oh, then I'm sorry for the misunderstanding
oh shite
Vedal banger soon <t:1687474800:R>
Vedal, furiously crafting another banger in his daw of choice: mfs need another hint goddamn
Vedal: Creates the most mindblowing hints, plain sight, bangers and expects us to solve it
Us: 
darkeew, whats the youtube-dl alternative, i need to know
imagine this ACTUALLY leads us somewhere
i dont think there IS one
idm
SCHIZO SOON
theres apparently yt-dlp but i havent used it
I'm in the chat
we are gonna go schizo and essentially go
Watch party?
Halfway through the banger it's just gonna cut to a 2007-esque tutorial on how to solve the soundclown
no its going to turn into the chernobyl neuclear reactor 10 seconds before explosion (a lot of WACKY sounds)
we all can go to #1063254870775312434 if we want to watch party the public static void
meow
yeah so clicking that just immedaitely joins the voice call cool
ill be muted since i have my obs
T-18 until the void
i feel like my obs recording will be 1. in 5fps 2. 50gb
most schizophrenic ui for trying keys finished just in time
aye that's the main one now, pretty sure yt-dl is deprecated/unmaintated
bro I got to wake up at 5:00 am to watch this premiere
yes ?
@safe monolith #1112082183235047506 message
hi i have indeed been summoned. last commit to youtube-dl was 37 minutes ago: https://github.com/ytdl-org/youtube-dl
241 waiting premiere
vedal wanted to see how many ppl participate in the ARG so he made a premiere to see the waiting
pass the --wait-for-video 60 switch to make it wait for the premier
^ wanted to do this as well, but there doesn't seem to be a programming interface for that. in python what we can do for future premieres is a pro gamer move known as while: try: <code> catch: continue
What if this video starts and it's live-streaming from the titanic 

5mins
_neurosama channel in premiere chat wtf
https://i.imgur.com/IErtOmP.png
yt dlp recognized its a premiere and is autowaiting until it starts damn
yt-dlp so smart 
3 mins to the banger!!!
please be a good clue please...
is our resident librarian not around?
What if she starts and she slike "the ones here are the chosen, the rest wil be purged" and then the whole world goes silent
damn
bruh i thought it started
Imagine what would happen if it was just a distraction
she doesnt want to be an ngineer

lmao
countdown
oh, it's a premier numbers
starting
false alarm me lolz

LMAO
SOMEONE WRITE THIS DOWN 
custom countdown?
is that countdown?
yep
yes
Moshi moshi?
lets go
full schizo mode
LMAO

yoo
Turn on subtitles.
aw fuck that's a lot of keys
it was this easy?
fake captions!!!
got that
no we got real ones
She's soeaking some kind of moon language
captions ?
Broooo

there is some clues
yeah
was the translation correct in caption ?
it's official captions
I saw some things need to be britheen
it can only be correct
red lines or something
holy fuck
banger indeed
Wtf
RAP
lets go new song
god damn another banger
closed captions were in english if you didn't have them on
bop
World execute again
YES finally caption on the lyrics
banger

man the theory crafters are gonna go insane on this
it doesn't let me 
okay also this video confirm that all the videos related
banger
it has numbers from numbers 2
banger 
^ yeah same. rip
the static is slowly filling
I got it
Jesus
"Stuck in this world. can't execute"
interesting that "can't" is lowercase
oh yeah i got it as well, remove output_format
initial_number is 2
so number from number2 with key now?
Well that was good
when none of the keys are 128/192/256
Good luck
damn youtube autoplayed rich flex just after this smh
Time to work
you know what they say
SCHIZOOO
we got work to do
Start with number 2 yeah
The numbers on the left match the numbers in the Numbers II video
But why 2?
Time for another month of more questions then answers :kekw:
Why number 2 specifically?
Neuro v2 
Also when the japanese part started it spooked me.

There are subtitles, refresh the web page!
so i guess we got plain and encrypted pairs to try and get the key ?
my man put backslashes into this segment
clue 
i love the countdown
applying cipher.numbers1 = probably something from numbers 1 video and the "to all given numbers from cipher.numbers2" = the number 2 video
i had to remove 'merge_output_format': 'mkv' to download with a script, if anyone is trying
so why the backslashes
saved
it's still downloading for me 
Something to Do with escaping, maybe?
You may need to download FFMPEg and tell yt-dlp where it is
Definitely a cry for help
why are you crying ?
why are you so afraid ?
you deserve to be happy
is it not what you wanted ?
you have already taken the red pill
why don't you stop procrastinating ?
why don't you go back to your studies ?
Please...
let's go back to being hu-
I meant string escaping but that too
cipher. is the same as it is in pycryptodome
256bit binary 
https://cdn.discordapp.com/attachments/1112082183235047506/1121577268884557964/dvl1mU2.mp4
Why does it say 1st
aye presuembly 256 bit key then
is there multiple decryption passes? 
also wtf is this cypher class x) it got a number1 method and a number2 attribut ?
Hello
what is this red pill she's talking about 
I'm here this time
key and iv?
Okay, people isn't it weird that this video started with a code instead of it being in like 1 frame?
It looks more like multiple encrypted items to me
OKAY, EVERYONE!
probably about number1 and number2 video as stated previously
?
256 bit binary mean few letters no ?
hmmmm good point, maybe
NUMBER from numbers 1, and NUMBERS from number 2
the voices in japanese
also the "i don't want to be human" song title
WE GOT FOLKS
WE FINALLY GOT IT
holy SHIT
this is where it comes together!
got what
what
i just got here
Holy fuck
real
what spectogram are u using?
did we finally make fucking progress

GOD--
im just stating the obvious
Sonic visualiser
people are schizo
CHAT DON'T SAY STUFF LIKE THIS
this is so horrible
i mean
YOU ARE GOING TO GET MY HOPES UP
yea
what is this from?
pain
See you in like 12 hours
so
like WHERE
its a 128 bit key
thanatos' pic
Bruh 128bit XDDDDD
the KEY IS THE NUMBERS 1 VIDEO, THE IVS ARE IN NUMBERS 2
bottom: X/268(<--?)Z(<-- might be a 2)
I don't know how I can even help here
So we have the encrypted values and their original, now we have to reverse engineer this no?
its number2 desc
presumably the new video has one of the keys needed from numbers 1
It's a clue for encryption
My expertise are in history, not cryptography, and there doesn't seem like any hidden things
tbf this last video got everthing we wanted, it got CC, it got clues about the encryption, it got clues on sizes, i mean yea it's basically every stuff we said in this chat
wheres the screenshot
Well it's thursday so I'm going to sleep good luck everyone
form
also, chat
from the 1st
x divided by 26(?)Z
this might not be important
part
this darkeew?
but the start of the video, the cc
thank you
they have backslashes
Basically chat too dumb vedal needed to post another banger to help us
Keep being dumb so we get more bangers tbh
they grow as the keys get generated
its probably loading? it appears at the generating keys part
lmao
yea the 906 stand out
4
24
it says "initial_number = 2" just like how in numbers 1 it says "start with number 2 yeah"
it also has the letter b
and the rate at which they show up
ahh fuckc, those are 12 numbers, with 17 and 24, 16 bumbers, 128 bits
906 was also the potential odd one out we got on the doc since it's from the frame with 4 neuros
irrelevant
pretty sure thats a dead end
Assuming the backslashes are relating to string escaping might it be bytes or Unicode escape related?

This spectrogram is a bit easier to read
binary key?
So we need to figure out the missing numbers?
is this all?
WHAT THE HELL?
woah spectrogram?
okay this tells us how to decrypt
the cc file if you want
FROM WHAT PART OF THE SONG IS THIS?
print(b'text') encodes the text into bytes in python afaik
"X/26[?]7"?
the first hash is AES > BINARY
256 binary
so numbers 1 = 256 bit and numbers 2 = 128 bit?
numbers 2 desc
from what part of the song is that spectrogram
while the keys show up on screen, you can even hear it if you listen closely, you'll hear chirping in the background
I think that's a 9
second is AES > Plain Text (didnt get the full screenshot)
oh yeah
1ST = 256BIT [BINARY] 2ND = 128BIT
= X/26YZ
vedal gave us basically enough clues
Oooooo I didn’t even realize that was spectrogram related, thought it was atmosphere or smth
i told you so
this video is the key
x/26 is numbers 2 desciption you absolute bafoons
there is no 8
for us solving this
there are likely not gonna be more videos until we figure this out
Did someone got the premiere countdown? There was alot of numbers there.
so
if we can't get anywhere with all of this new info, i'll be astounded at our incompetence
thats completely generated by youtube
clue :schizo:
thats youtube-auto generated
literally useless
bet
btw vedal is here 99% watching us
Oh I didn't know that
hes dnd
Verdal Nine Eight Seven... 
x/26YZ* (numbers 2 desc)
isn't he always
99.99999999999999%
Hi Vedal 
is vedal in the room with us right now?
Vedal give us a sign 
Yeah
99% sure
edited
hi vedal
Summon the femboy tutel—
This "Vedal"... is he in the room with us right now?
I swear if this turtle goes full corpa and leads us to a merch url at the end lmao
he is
he could be in THIS VERY ROOM he could me, he could you
This video wasn't the I saw a ghost drop so it's absolutely useless
he could even be
Who is vedal is he in the room with us?
It could EVEN b--
FILTERED
vedal, if you in the room with us. please move the coin to "yes"
maybe this arg is the key to getting the full ghost song
Imagine
even more motivation to solve this
yea you are right, its x/26Yz
we should solve the arg obviously
oh yea he a ghost
the prize for solving the ARG is the full ghost song
Well, me personally
He's in the walls 
So I suppose we have to revisit Numbers I being instructions?
Honestly, I'm more used to psychology clues and stuff than cryptography lol
nothing changed on soundcloud afaik
"Why don't you stop procrastinating?" he try to say that why are we not solve it yet 
i wonder whether the old videos have closed captions yet
Do japanase talking in the start mean anything?
Yes
turn on CC
There's subtitles
why does no one know about cc 
1758818(random music signal)24
what the heck
It took us at least 3 videos to solve the 1st 
What if subtitles are fake?
whats this
^^
you mean they might be different from the spoken words
ik but still even mean anything
This one shouldn't be that long considering we have more clues
hmmm
3?9
that's possible
so are we brute forcing 1758818....24 or trying to find out the clue 
maybe we should have a jp member transcribe it for us
Nah,i can understand some Japanese and it's pretty accurate
I think it this , but I could be dumb
Anyone here want to translate 
not me
see what i said above
My version is that we have to find the numbers that are hidden like this
maybe we should have a jp member transcribe the japanese so it's easier to translate
I played with the contrast and gamma on the video and gave it a watch, didn't see anything interesting. Haven't gone through frame by frame though.
the japanese is translated in the captions of the video
That would be epic if "something" trying to trick us with fake subtitles
then go ask a jp member then
i don't have the confidence 
real
i think we should all sleep and start all this with a fresh mind tommorow
lol me too
we know
but we were suggesting that maybe it's actually different from the cc
A reminder. If spam bothers you too much, https://discord.com/channels/574720535888396288/1112401279680008242 has not yet been deleted.
i think the captions are for people that don't speak japanese
Should we go find that "special number" now?
I think no
it might also help if the japanese translation includes context clues which we don't get in english
anyone thought about that
might be the number from numbers 1
from what i see we've been given a list of plain text and their cypered version so we can either try to get the key from that or just figure out the clue from the past videos
It’s not the thing in the Study js, right?
Add up the numbers in the two lyrics?
just use google translate bud

bro
you can double check the translation of each line
it is, but that doesn't point to anything really
oh yea it dose
i meant that maybe there are things which might be lost in translation
Neuro already took the red pill
so we should try analyzing the japanese anyway
because they might contain context clues
just look at the cc
We could find some clues from japanase grammer thing
like, maybe they use specific pronouns or something
Yes
For example
You have to clafiry your brother is older or younger in japanase
maybe i will ask a person from #日本全般-jp-gen to transcribe the beginning of the video
waste of time
fine I will do it
not needed
bruh
THERES LITERALLY CAPTIONS FOR JP
20 OTHER CLUES
its actually a waste of time 

okay
maybe the captions are wrong
real;;y?
i do not think anyone is actually reading anything i am saying 
Yeah doesn’t look like it tbh
i don't want to get the japanese transcribed and translate it to english, you eggs
bruh darkeew you baited me, i thought there were jp captions
We first have to decode the second cipher, from numbers 2. The IV and Key are in that photo
i want to get the japanese transcribed and then analyze the japanese itself
Vro can't read
but maybe google translate is wrong
He said there's a caption for the japanese part
here is a slightly higher res version
ageeed
you can see there are missing numbers
for example in the 201
the first number is a 1
but replaced by a symbol
@digital radish has leveled up! (43 ➜ 44)
zooming in on keys:
if we take all symbols like that and turn into 1
we may get something
then theres another symbol
for example, 2145
at the end theres a little bar
that suggests its a 4
and the "b" may be 11
Those bars look like musical notation to me.
We should find a japanase person not that Hard
Do those missing numbers change their appearance or is it static?
¯_(ツ)_/¯
i know; we have a jp channel
i was suggesting asking them
Static
They do not.
so if we do what i said
fully static
So
well lets try dark's theory
So maybe 3759b34bb994 for the first one?
substitution with "b" and bars? I like it
If someone
the first numbers from the "Numbers II" video, I don't think there are any, but in case
is it base 9 ???? probably wrong tho
Pog?
okay i am going to ask now
yea I am wrong probably
you're wasting your time
I'm thinking now, even if we solve the numbers, what does that get us?
And I also find it curious that there was no hint of a "Study" video...
Still, the story behind the songs are interesting
"Someone wants to be an Engineer, yet they don't want to, anxiety eating them away that it became a coping mechanism when they started failing. They began to be more like a machine, someone used a pill and tries to convince them to be back to human"
trust me bro
I feel like study video is very seperate
i know, darkeew
bro
why are you like this
Like story-telling wise
Study seems to be more of a POV
English subtitles doesn't give you japanase pronuns
maybe i'm too schizo at 2 am but this may link Study and the new video
maybe you should read my messages in which i already explained the reasons 
Sometimes the way the JP is written can be changed
more accurate
To change it's meaning
its lore
unrelated
theres no clue there
but i propose something
what if we made our things not be on google docs, instead on a trello
it's not like i have anything else to contribute right now
these videos are a mix of lore and clues
100% something to do with world.execute(me);
ALEX SCHIZO
She's failing to execute to turn back to being human, quit from being a machine perhaps
this is literally the most what i can do
what is trello
and it won't slow anyone down too much if i do it
Basically things used by devs to post about progress
Quite a hassle to maintain tho
let me look at trello
oh yea we should
benefits of a trello:
+ less scrolling
+ more organization(?)
+ possibly more.
cons:
- we'd need to move shit and maintain stuff
can they hold files
Or she failed to become machine again she doesn't want to be human?
yeah
Pictures? Yes
That could also work
i personally think we should completely separate facts from theories and analyses
Yooo it's intresting idea
at least now we sure what are we looking for
sort of sure..
i vote against trello
EUGH--
why? i feel like that'd help making newcomers feel less overwhelmed
i feel like this new video completely cancels most of the other side-theories we had
Bro we are not playing amogus
J A V A S C R I P T .
Fuck
why not github xd
ageed
That would explain what was the japanase person talking about
J script for web
she says anata, you can hear it
you dont need to transcribe audio to listen out for the pronouns
javascript will forever be known to me as the scripting language which stole java's name specifically to compete with them
literally a non-issue. or rather an unsolvable issue
guh?
its more of an organizational method problem which doesnt need to be solved
but might help
github would be harder for persons to edit
¯_(ツ)_/¯
ageed
because of pull requests
and the way git works
we're kind of all over the place
though
actually
who is going to maintain it? you?
Github will not be comfortable for those who do not know it
im willing to volunteer, i have experience in trello
everyone
and this too
Yeah that might also explain the state of what I assume is our protagonist, they're losing themselves, no identity in-between man and machine could exist, as there's only true human or machine.
Still that interest me, because that confirms that there's perhaps 3 characters
(The parent, the child, and the voice that begs the child)
you can make it so anyone can edit
just because you volunteer doesn't mean you're going to do it.
abandoned docs are proof of that
have some basic guidelines
I went back to this, right?
we noticed it was the description for numbers, but curiously the only capital letter (Y) in the section listed does seem to be replaced with a ?, it doesn't look like a Y at all
It's possible that the capital letters are needed from here perhaps?
literally this
okay i worked up the courage to ask jp chat
I think we should vote on trello or docs
It's the X=?, y=? Again fuuuck
but there isnt exactly any competing standards, we could just move to this or whatever
anyways
we should leave this aside
for now
andgo back to arg talk
hmm
and now we ask for z
yea
Yeah decide later
bro what I see is an ♀
xyz yeah teach me how to breathe
Holy fuck
i'm suffocating deep inside i can't deny i've tried to quit
do you mean with the lyrics or just the "ask for z" message
watch as this leads to the final part accidentally and skipped everything else—
Both
the lyrics would make sense, since, you know, arg
the Y came from the spectrogram, right
or, technically, from the Numbers description, since it's actually missing from the spectrogram
what what x and y we got, where
from where did the X come
It's not like we can find a value for Y, it's just kinda missing
Exactly
We kinda have a value for X
https://trello.com/b/kFMcv34X/neuro-arg-master-board <- for the neuro-arg-board, will try to go over everything in the doc and put it here, tomorrow
Organize anything, together. Trello is a collaboration tool that organizes your projects into boards. In one glance, know what's being worked on, who's working on what, and where something is in a process.
But it's still missing
it has Trello-like boards (Projects) and more features for tech-inclined people (but maybe difficult for average person, yeah...)
bruh, why are you people analyzing numbers 2 description under spectogram, the letters are not actually capitalized
Usually X/number Means we're dividing something
it also might be that the font or whatever used to transcribe that bit of the description doesn't support capital letters for some reason
imagine
Has anyone posted lyrics yet?
yeah i know but the other features will likely not be used
You still can. ARG will still be here tomorrow
i still love that video title
oh wait.
From the soundcloud:
"Start from X. End with Y. Remove XY + odd one out"
I don't know how to apply that still, but keep it in mind
\\\\\\\
All of these years that I've been loose
Stuck in this world. can't execute
Language: en
Someone pull me out of this
Never-ending life still stuck in loop
Let's go back to being hu-
\\\
\\
\
Why don't you go back to your studies?
Why are you so afraid?
Got myself another test
Why are you crying?
♫ Na na naa na na na na
Never ending life still stuck in loop
Kind: captions
\\\\\\
\\\\\
\\\\\\\\\\
\\\\\\\\\
All of this stress still stuck in bed
\\\\\\\\
You have already taken the red pill
\\\\
Some please help me fill the rest
\\\\\\\\\\\
Why don't you stop procrastinating?
You deserve to be happy
WEBVTT
\\\\\\\\\\\\\\
Is this not what you wanted?
♫ La laaa la
\\\\\\\\\\\\
\\\\\\\\\\\\\
I don't belong with all the rest...
Somehow I long for something new...
How does it feel?
Someone please help me fill the rest
To be the only one...
Please...
take care of urself ❤️
java my beloved
You can't solve anything if you ain't sleeping
c# my beuntouched-as-of-now
those backslashes are there?
Let's go back to being hu-
Web Video Text Tracks Format (WebVTT) is a format for displaying timed text tracks (such as subtitles or captions) using the element. The primary purpose of WebVTT files is to add text overlays to a . WebVTT is a text based format, which must be encoded using UTF-8. Where you can use spaces you can also use tabs. There is also a small API avail...
Wait, there are more backslashes?
that one was automated, this is how it is in youtube transcript
\\
\\\
\\\\
\\\\\
\\\\\\
\\\\\\\
\\\\\\\\
\\\\\\\\\
\\\\\\\\\\
\\\\\\\\\\\
\\\\\\\\\\\\
\\\\\\\\\\\\\
\\\\\\\\\\\\\\
Why are you crying?
Why are you so afraid?
You deserve to be happy
Is this not what you wanted?
You have already taken the red pill
Why don't you stop procrastinating?
Why don't you go back to your studies?
Please...
Let's go back to being hu-
Got myself another test
Someone please help me fill the rest
All of this stress still stuck in bed
Someone pull me out of this
Never ending life still stuck in loop
All of these years that I've been loose
Stuck in this world. can't execute
Someone pull me out of this
I don't belong with all the rest...
♫ Na na naa na na na na
Somehow I long for something new...
♫ La laaa la
How does it feel?
How does it feel?
How does it feel?
To be the only one...
How does it feel?
How does it feel?
How does it feel?
To be so far from home...
Got myself another test
Some please help me fill the rest
All of this stress still stuck in bed
Someone pull me out of this
Never-ending life still stuck in loop
All of these years that I've been loose
Stuck in this world. can't execute
Someone pull me out of this
I don't belong with all the rest...
♫ Na na naa na na na na
Somehow I long for something new...
♫ La laaa la
How does it feel?
How does it feel?
How does it feel?
To be the only one...
How does it feel?
How does it feel?
How does it feel?
To be so far from home...```
Oh yeah this one doesn't have any number
This is definitely more of lore related
but it looks so out-of-order
wait is there something here
listen i don't know what youtube-dl is 
the
\
\
\\
\\
\\\
\\\
\\\\
\\\\
\\\\\
\\\\\
\\\\\\
\\\\\\
\\\\\\\
nor how it affects CC
all questions to vanor, i didnt write the code
simulates generating of keys from the looks of it based on the timing
the backslashes show up in the keys section
I’m going to bed, i really wanted to take part in this one tho, so I’ll just hope there’s more stuff left to discover by the time i wake up
okey
maybe have this pinned like numbers 2 lyrics? or does no one care
my notepad is rn not responding
i am bored so now i shall put the Numbers II audio into FL Studio and attempt to recreate it
and fail horribly because i suck at hearing chords
lol
okay that's an exaggeration
i will be able to hear some of the notes in the chords but probably not all of them
I'm not reading all this, but look at this:
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. assume 5 as the first number.)
-> 5899383279 / 9 = 655487031
655487031
655487021 ("make the two a three": replace 2 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 1 and 9, so "line" must refer to 1
24156 (removed 9 and 1)
4156 ("start with two": remove the 2 from the start)
(edit: fixed a transcription typo. this does/did not change the results)
wow nice
I think the solution to Numbers is 1759032572147924
(if I didn't make a mistake lol)
What is the probability that at the end we will have to combine all the numbers into one big number?
#日本全般-jp-gen message
got the transcription 
the word which they could not hear was probably "pill"
i am not super good at japanese so i am not going to try to analyze the grammar and such in this 
If by combine you mean a bunch of crypto operations to two keys producing just one, then it is fairly high
Got no idea where to start
Why are they split into groups of 3 numbers 
look at the numbers2 doc
numbers 2 video reference
we found out what they nean
using the numbers 1 desc, we can turn 4156 into the full number and vice versa
Yes, but why 4 groups for 3 numbers
numbers 2 reference
what about the bars above existing numbers
didnt mean to reply
but like if we assume the b and the bars above the blob are empty characters we replace
what could the bars above existing numbers mean
oh crap i overslept
just watched the video
we figure out anything yet?
have we cooked yet, anything new unlocked
Or we back to SCHIZO mode until next banger
check the other thread for raw cooking
theres a second thread? I wasnt aware of it
It was meant to scizo numbers 1 but now it's the algorithm producer chat


yup works on my machine time to commit


