Hey folks,
Been a while since I've posted anything on this but I wanted to share my thoughts on what my approach would be to decode it. First of all let me get a few things out of the way because there's been some speculation that there may not be any pattern to it at all and it's completely random. Well, I'm not buying that. First of all we have the hint from one of the developers that we should listen to it (which just seems too arbitrary to put out there without a reason) but more than that, I've found patterns in the data which just wouldn't pop-up as often as they do if it was entirely random.
UNKNOWN ARTEFACT: More audio and giveaway
UNKNOWN ARTEFACT: Text translation pattern scanner
After comparing with random number generators you just don't get sections stringing together like that as often as they do. Anyway, at first I figured one of the major problems with decoding this thing is that not enough of it was recorded. Without having a complete recording which repeats from start to end over and over I figured, why even try? Take a segment out of a bit of binary and it can totally screw up the data - and I still think if we had a ridiculous amount of recordings (hours and hours) then we probably would see it repeat enough to be obvious what was going on - however now I'm thinking it might not be that time consuming.
First of all, I'm not convinced all the weird noises in the background mean anything. These kinds of ambient sounds occur everywhere in the game and vary depending on what you're looking at or where you are - plus they're too subtle to clearly determine what's what within them. The big thing with the UA's is the binary pulses. They're clear, loud and obvious - as you'd expect them to be if there was a message within them. In most games, developers give the players clues to solutions to problems - I mean you don't have to click on every single pixel in a point-and-click adventure game to find that missing object - there's a hint or an image to guide you.
So sticking with the idea that it's just the binary we should be looking at, I started thinking that since we get repeating sections but not an entire loop, perhaps the message is fragmented and we have to piece it together before we can read it. I thought maybe half of the binary could be an indication of where each piece goes or something, and the other half was the message itself. Perhaps 2 or 3 bits are a marker and the other 3 or 4 bits are the data...and these sections come in different orders. Although I haven't looked much at it nothing's jumping out at me however so...here's another thought and basically the point of this whole post so far: Maybe we've got two or more loops of binary playing together and they're different lengths - all we have to do is figure out where each one starts and finishes.
If I try to illustrate, imagine you have this: 0111, 001100 and 11010101. If I write them all down together, the pattern always changes but the loops themselves always stay the same. Check it out:
001
101
110
111
000
101
100
101
011
111
100
101
000
101
110
111
001
101
100
101
010
111
100
101
As you can see, each pattern repeats independently but the entire thing as a whole almost never looks the same anywhere. I just wanted to suggest that maybe we already have the entire message in the few hours of recordings available (possibly even several times) but we just haven't noticed it.
Been a while since I've posted anything on this but I wanted to share my thoughts on what my approach would be to decode it. First of all let me get a few things out of the way because there's been some speculation that there may not be any pattern to it at all and it's completely random. Well, I'm not buying that. First of all we have the hint from one of the developers that we should listen to it (which just seems too arbitrary to put out there without a reason) but more than that, I've found patterns in the data which just wouldn't pop-up as often as they do if it was entirely random.
UNKNOWN ARTEFACT: More audio and giveaway
UNKNOWN ARTEFACT: Text translation pattern scanner
After comparing with random number generators you just don't get sections stringing together like that as often as they do. Anyway, at first I figured one of the major problems with decoding this thing is that not enough of it was recorded. Without having a complete recording which repeats from start to end over and over I figured, why even try? Take a segment out of a bit of binary and it can totally screw up the data - and I still think if we had a ridiculous amount of recordings (hours and hours) then we probably would see it repeat enough to be obvious what was going on - however now I'm thinking it might not be that time consuming.
First of all, I'm not convinced all the weird noises in the background mean anything. These kinds of ambient sounds occur everywhere in the game and vary depending on what you're looking at or where you are - plus they're too subtle to clearly determine what's what within them. The big thing with the UA's is the binary pulses. They're clear, loud and obvious - as you'd expect them to be if there was a message within them. In most games, developers give the players clues to solutions to problems - I mean you don't have to click on every single pixel in a point-and-click adventure game to find that missing object - there's a hint or an image to guide you.
So sticking with the idea that it's just the binary we should be looking at, I started thinking that since we get repeating sections but not an entire loop, perhaps the message is fragmented and we have to piece it together before we can read it. I thought maybe half of the binary could be an indication of where each piece goes or something, and the other half was the message itself. Perhaps 2 or 3 bits are a marker and the other 3 or 4 bits are the data...and these sections come in different orders. Although I haven't looked much at it nothing's jumping out at me however so...here's another thought and basically the point of this whole post so far: Maybe we've got two or more loops of binary playing together and they're different lengths - all we have to do is figure out where each one starts and finishes.
If I try to illustrate, imagine you have this: 0111, 001100 and 11010101. If I write them all down together, the pattern always changes but the loops themselves always stay the same. Check it out:
001
101
110
111
000
101
100
101
011
111
100
101
000
101
110
111
001
101
100
101
010
111
100
101
As you can see, each pattern repeats independently but the entire thing as a whole almost never looks the same anywhere. I just wanted to suggest that maybe we already have the entire message in the few hours of recordings available (possibly even several times) but we just haven't noticed it.
Last edited:
