UNKNOWN ARTEFACT: Sound Engineers, apply here!

Here's how I understand it.

After analyzing the involved sounds. There are three different kinds of "Morse" signals. (After the "Whale" sound) arranged in different sequences. The first "High" signal has a fast attack. The middle Low frequency " Purr", and the third High signal with a "low" attack. These signals are arranged in groups of up to 7 in between the "Whale" sounds of which three types stand out. The first and second are almost identical. the third has a long tail off.
Regarding the "Morse" signals. I would imagine the High attack means a leniency towards the left Negative. The middle F=0, and the High FQ with a right sided attack as Positive.

I had an idea then that people are looking for a 0 point. Sol, or Sag A. It occurred to me that there really is no reference apart from the whole in Galactic standards. If you look at the Gal map, you see three arms on one side, a Middle bar, and Three arms on the other side. That is seven separate locations. The problem I have with this idea is that we would need a central reference (N,S,E,W) as well as an angle to find the location. Unless there is an eight bar indicating depth or height.
 
Replying here just to give some input. BBC Micro could save to an analog tape deck. Not sure of the format but this could be a recording of a tape being loaded into binary. what happens if you make it into a string and assume it was 7bit with either even or odd parity. Which is an option of data transfer back then.

From wikipedia

Tape interface (with a relay operated motor control, controlled via 2 pins on a circular 7-pin DIN connector), using the CUTS [Computer Users' Tape Standard] variation of the Kansas City standard data encoding scheme operating at 1200 or 300 baud

Maybe not by looking at the Kansas City standard page.
 
Sounds just like when a capital ship enters but with a Thargoidian twist. I think its that simple. "Hey guys listen to this!". They are coming!
|
Probably no tin foil moments here, but hey i cant argue with the fun of it. :D
 
From the reddit thread, someone has given us a lead and direction.

Hello Everyone (and to the OP, I was just about to PM you) -

That was my post... I was using DigitalScream's long recording, which includes 2 very similar sections. I haven't yet posted another message, but suffice it to say the the additional data I have seen, leads me to believe I was incorrect in some of my assessments. I do however still believe that there is a binary signal as the "principal" element of the audio.

Incidentally, I'm really happy to see this thread -- I just finished a PM to DigitalScream, asking if we could organize some kind of audio collection effort... and here it is!

As for why I think my previous theories don't hold up at present:
  • Sample length: I now realize that virtually all the samples I've listened to are incomplete, making any assumption about "length" packets, or number of byte runs, questionable. What I don't understand is: why do the samples end early? Are the artifacts about to decay or something ?
  • 'Word' size: The additional samples appear to indicate (clearly, with the ua_2 sample) some busts with numerous 6-bit words. Thus, any universal assumptions about 7-bit alignment are questionable. Either the words can vary, or some bits are inaudible -- I hope this isn't the case, as adding signal 'corruption' into the mix will make this extremely difficult to solve
  • Incomplete headers: the headers seem to be 5-bits, but some samples begin partially into these bursts. It would be good to being recording during the deployment of the UA's themselves.

This brings me to my PM: We Need Data! I can think of three principal reasons:
  • Duplication of results: Is the same data EVER repeated? Again I hope so, and finding a signal in more than one sample will be a huge step forward in decryption
  • Filling in gaps: Similar to above, if some of the variations are random, we can parity check other binary transcriptions
  • Signal Processing: It provides more data to the guys doing signal processing to the audio itself, in case there's more there.

I have a few scripts I've written to aid in the organization and processing of the presumed binary data. Is anyone else interested in working with these? They're super simple to use, and only require installing Python, and basic knowledge of the command line

- - - Updated - - -

Forgot to mention: I do strongly agree that we need to organize for as much collection as possible, organized in a single location. From there, we can begin associating various supplements, like the transcriptions
 
Last edited:
From the reddit thread, someone has given us a lead and direction.

They come across as "Nyaah! I know something you don't know and I feel all superior about it"
There's very little of substance to the posting. Sure, we should bear in mind the possibility of a voice in the sound, just like we should bear in mind the possibility of a coded sequence in the sound.
 
Possibility? There is a coded sequence in the sound, that is indisputable.

Whether it has meaning or is random, and whether we can decipher it, are entirely separate matters.
 
They come across as "Nyaah! I know something you don't know and I feel all superior about it"
There's very little of substance to the posting. Sure, we should bear in mind the possibility of a voice in the sound, just like we should bear in mind the possibility of a coded sequence in the sound.

For the entire UA hunt, reddit has been a source of intentional misdirection (hoaxes) and noise while these forums have been the source of (every?) actual breakthrough (and noise :D ).

I love the elite subreddits, but in this one instance I'm subscribing to the idea that following the analysis here and ignoring reddit "breakthroughs" is sufficient for the task, and more than that, probably also more efficient too. :)
 
Last edited:
So, instead of speeding up the audio to listen for Highs and Lows, (which I believe is clouded by the erratic 'heart beat' sounds it makes, causing there to be no discernible pattern) I listened to it at plain old 1X, and I hear a pattern of 7 notes. First four (ending with the iconic foghorn note), then three. These repeat, without fail, every time. AND they coincide with when the UA 'lights up'.
Now, I'm not fully convinced that the Thargoids 'want' us to find them, because if they knew our culture they would just emit a message in english (for example), but for now let's assume that for simplicity, we can use all our human tools to decode it.
I took the seven notes and corresponded them to notes on a keyboard.


The first is D (27) - A (34) - D (27) - D (3)

The second is F (30) - C# (2) - D# (28)
I'm trained in music, but by no means a tone expert so I would love for someone to check me. This results in seven alpha notes, or 12 numeric digits. These twelve digits could be used to make a 3-axis coordinate system, with 4 digits each. You would result in (27.34, 27.33, 02.28)
Just one of the many ways to interpret the sound, but I believe looking at the notes themselves is more reliable than the binary system of highs and lows, because there is a definitive pattern, and it's not just high and low, there are multiple notes being used.

Someone on Reddit worked out the tones of an artefact were C A# G E the first thing that poked out to me was a system name of origin. For example with this code it could be "CA GE....." the others spell the origin system of the artefacts maybe leading to powerplay that will release something within that system. This could co-inside with your theory.
 
Someone on Reddit worked out the tones of an artefact were C A# G E the first thing that poked out to me was a system name of origin. For example with this code it could be "CA GE....." the others spell the origin system of the artefacts maybe leading to powerplay that will release something within that system. This could co-inside with your theory.


The cage....

Do NOT go there....


one bleep for yes
two bleeps for no
 
Hi All,

After sitting in a waiting room pretty much all day pondering and listening to the audio in any which way one can, I think we may be looking a little to deep into this. For me the base 7 theory does not stand as as others have said the sequence breaks.

Stating this and remembering what FD said, could this be some sort of tracker beacon instead of being a coded message? If the audio pattern changes when jettisoned and targeted in other system then we really need to concentrate on the changed segment only - this could turn out to be a very simple hot/cold direction/location finder where the audio signals change the closer (or farther) one is from the target.
 
Hi All,

After sitting in a waiting room pretty much all day pondering and listening to the audio in any which way one can, I think we may be looking a little to deep into this. For me the base 7 theory does not stand as as others have said the sequence breaks.

I would think of it as a bit stream in whole and ignore the breaks. The breaks could just be a pause or emptiness for a protocol.
 
Structure is the opposite of random. There is clear structure here: A tone plays only once or twice, then the other tone plays only once or twice. These may be punctuated every 5-7 tones.
For a real-world example of this structure being used to transmit data, look at an ITF barcode.

(If you were to transcribe actual ITF into UA tones, I think you'd get groupings of 10-14 tones in length, double the 5-7 length of the UA groupings. Ie. I'm not suggesting the UA is talking in ITF, just pointing out that this signal is slapping you in the face with clear structure. :)

100% Agree with this. The lack of a 3-run of 1 or 0's, contrary to what some have said, is a *very* clear indication that the signal contents aren't noise. Statistically white noise would have to produce runs of 3 or more.

The "barcode" theory is something I've been looking at since yesterday, and it a potential lead. Here's a simple example of barcode "no 3's" encoding:
http://en.wikipedia.org/wiki/Interleaved_2_of_5

- - - Updated - - -

For the entire UA hunt, reddit has been a source of intentional misdirection (hoaxes) and noise while these forums have been the source of (every?) actual breakthrough (and noise :D ).

I love the elite subreddits, but in this one instance I'm subscribing to the idea that following the analysis here and ignoring reddit "breakthroughs" is sufficient for the task, and more than that, probably also more efficient too. :)

Maximizing both is in the best interest of maximizing the effort of the community, IMO
 
Last edited:
The "barcode" theory is something I've been looking at since yesterday, and it a potential lead. Here's a simple example of barcode "no 3's" encoding

People have also mentioned that the structure is likewise similar to sending data by modem or saving data to audio cassette tape.
I like this angle because the original Elite was sold on audio cassette tapes, which adds an element of nostalgia-easter-egg :)

But the signal had better not decode to LOAD "ELITE" :D
 
Last edited:
There may be a voice in the Ancient artifact transmission. Here's how I found it:


Tools: Audacity
Sample used - Digiscream's audio http://www.digitalscream.org.uk/audio/unknown_artifact_audio.mp3

Method:


1) cut off the in-game audio (cargo scoop) at the end of the sample
2) Normalised remaining sample to 0DB
3) Pitch shift up by two oactaves


4) using audactity, I duplicated a portion of the background hum, and used this as a sample in the "Noise reduction" tool. The same effect may be done better by using a band pass filter. This was a quick and dirty attempt, with a little time we can clear this up a lot.


5) And here's the result:
IEJKPD3.png

CAUTION: turn down your volume :)

UPDATE:
Here's just the relevant small Snippet, with EQ boost in the 1-5k range:
http://vocaroo.com/i/s1DcReyHZVyf

And here's the full repeating pattern pictured above.
http://vocaroo.com/i/s0hIXtltcwMq


There is (*to my ears) a human (???) voice after each burst of whale song. It sounds like a numbers station.


Someone with better audio skill than I should be able to process the audio with more clarity. This was inspired by the recent reddit post, which may not have been a hoax after all :)

Edit: Here's the reddit post that inspired the process. I was attempting to see if this post was a hoax or not: http://www.reddit.com/r/EliteDangerous/comments/353aao/ive_managed_to_read_unknown_artefact_message/
 
Last edited:
Structure is the opposite of random. There is clear structure here: A tone plays only once or twice, then the other tone plays only once or twice. These may be punctuated every 5-7 tones.
For a real-world example of this structure being used to transmit data, look at an ITF barcode.

(If you were to transcribe actual ITF into UA tones, I think you'd get groupings of 10-14 tones in length, double the 5-7 length of the UA groupings. Ie. I'm not suggesting the UA is talking in ITF, just pointing out that this signal is slapping you in the face with clear structure. :)

So far, it never repeats. What does that mean? It means there is no message to be decoded.
 
There is human voice after each burst of whale song. It sounds like a numbers station.
Someone with better audio skill than I should be able to process the audio with more clarity. This was inspired by the recent reddit post, which may not have been a hoax after all :)

Very interesting. Thanks for the effort and providing your method.
 
There may be a voice in the Ancient artifact transmission. Here's how I found it:


Tools: Audacity
Sample used - Digiscream's audio http://www.digitalscream.org.uk/audio/unknown_artifact_audio.mp3

Method:


1) cut off the in-game audio (cargo scoop) at the end of the sample
2) Normalised remaining sample to 0DB
3) Pitch shift up by two oactaves


4) using audactity, I duplicated a portion of the background hum, and used this as a sample in the "Noise reduction" tool. The same effect may be done better by using a band pass filter. This was a quick and dirty attempt, with a little time we can clear this up a lot.


5) And here's the result:

CAUTION: turn down your volume :)

http://vocaroo.com/i/s0hIXtltcwMq


There is human voice after each burst of whale song. It sounds like a numbers station.


Someone with better audio skill than I should be able to process the audio with more clarity. This was inspired by the recent reddit post, which may not have been a hoax after all :)

Thanks. Will definetly have nightmares this night.
 
Back
Top Bottom