Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[bertix]

Sorry for double post but it may be an http://en.wikipedia.org/wiki/Arecibo_message

[Sirian]

Now only the 'AT 5.534417156693186' part must be understood.

5.534417156693186 seems to be the pulsar frequency. (5.534417156693186 Hz)

I didnt discover it, i just checked the 0x10c forums and someone pointed out that there are pulsars with such frequencies

That would be the reason for the "redundancy" bertix is talking about, basically we have 4 samples per second of a 5.5344... Hz signal

So thinking about it, that means we have a very low resolution of this signal, so low that maybe the idea is to recreate the pulsar signal, supposing that it is always the same, so that would mean this signal lasts 0.1806874999999999918013046875 seconds (1/5.534417156693186) and the values we have are dots of the signal graph, so to get back the real signal you just need to make a graph of the 0.1806875 sec sequence and fill the gaps in the data with successive iterations of the sequence. Then you get the message, whatever it may be.

If i'm not clear to you, think of it as a distress signal of some sort, where a message is repeated continuously. Let's imagine that the message is "please help me". But now imagine that this message is broadcasted on fast forward, and we pick only bits and pieces every time we hear it, so we only get letters like "p..a.. h..p..e" ".l..s..e.. ." etc... but eventually if we hear it often enough and put all the pieces together, we can have the full message. Here the message is repeated 5.5344... times per second and we get a bit of it every 0.25 second.

That's my theory anyway.

Edit : thinking about it, it makes sense that we have to make things fit in exactly 0.1806875 seconds, because the "5.5..." can only be a frequency, seeing as it is just precise enough that it's multiplicative inverse (the duration of a cycle) is almost 0.1806875 but since the frequency number would be too long if he tried to give it perfect accuracy, we have instead this almost correct value, which indicates that the "5.5..." number has been rounded from 1/0.1806875.

[bertix]

I would have to disagree with that,that frequency is the sample rate, not the frequency of any of the data the highest frequency we can sample in the data is half that Fast Fourier Transforms would help...

[Svampapa]

16 sets of number to each row...?

Maybe I'm throwing red herrings around, my skills really lie elsewhere.

But I just can't resist these types of things. Happy cracking, I'll be sure to watch you guys scratching your heads in here.

[bertix]

Made progress, the last stream has 3 different frequencies , high low and inverse, and a frequency lowering every sec

[Fiskav]

.

[Sirian]

I would have to disagree with that,that frequency is the sample rate, not the frequency of any of the data the highest frequency we can sample in the data is half that Fast Fourier Transforms would help...

I'll be applying my theory on a large chunk of data that i've got from running the ARG for a while. 0.1806875 is 0.0000625* 2891 so i'm guessing the message is made of 2891 samples. Since the game uses 16bit computers, i'm also guessing that each sample is made of 4 hex numbers ( like "FD34" for instance). Basically my idea is that we are sampling at 4hz a pulse/message that is repeated at a 5.5344... Hz frequency. So our sampling rate is asynchronous with the pulse which means we eventually read the whole message but all scrambled. I could be wrong in the sample duration and size but i'm going by my intuition, we'll see if i'm right.

Edit : By the way, my assumption is that there is only a small message scrambled all over a large stream, whereas you seem to assume that the stream is a long message.

[bertix]

applying my theory on the frequency's i could get that result, if this concurs with any of your research please do say.

[Sirian]

someone say this in English, please.

Imagine there is a message made of 2891 "words". The web page give a bunch of those words, but they're all scrambled. The method used to scramble them is to count the list of words going back to the beggining when you reach the end (the 2891th word) and pick the word you get when you reach the count of 4000. You then display this word and resume counting to pick the next word.

Here a word is actually a sequence of 4 hexadecimal numbers, which equals 16 bits, a common format in computer files.

So put the words back in the right order and you get the message (or file).

Edit: i may have gotten something wrong when i said this thing about counting to 4000, i'll have to double check my reasoning.
Edit2: nope 4000 sounds right. Others on the 0x10c forums have reported a 4000 periodicity so maybe i'm onto something.

[LoSboccacc]

it changes at every refresh; anyone checked if it is a scrolling sequence?

[bertix]

It does but its just huge dumps of cycling data...

[Girlinhat]

This is extremely interesting.  I love how everyone is applying some serious science to this.  Anyone made any actual ground on figuring out what it is?

[Brotato]

My science-y skills lie elsewhere.  Wish I could help you guys :/

[Eagle_eye]

I'm getting a message telling me there's not enough power, followed by a percentage between 63 and 64

[Matz05]

Apparently, the "power" is based on the number of people connecting. more is better.