This is real - Emily Audio

[toddumptious]

@sheralmyst

Am I seeing this correctly?

31 December 1969

That is the day before

This date showed up in the beginning of discovering the ARG too, it actually was in the 1970 log for me the first time if I remember correctly, there’s some weirdness going on there since it’s at midnight and sometimes displayed as 31 dec 1969. I think I may have archived this in my original post from the first day when cataloguing the entries. Lemme go have a look

edit:

I had to take a picture with phone because desktop screenshot wouldn’t capture it showing the link url down bottom right on hover, to show its from the 1970 entry log.

this is from an early post from first day found here

Project skyscraper / no man's sky arg

this feels VERY waking titan. If this is fan made, they certainly nailed the vibe. Hello World So it’s referencing memory blocs, and theres another area that shows memory bloc status as being at 0% It seems to track how many people are actively viewing the site, and the security log updates seem to almost be in response to user activity or fudged a little to make it appear as so.

[vector_cmdr]

Hm, now you mention it, this is anomalous actually.
1969-12-31T23:00:12+00:00 for the meta tag, and if OpenGraph was mangling it, you’d kind of expect it to not be so specific and be a 0000-00-00 00:00:00 default… and a timezone conversion error in the OG plugin on WP seems unlikely to render that as well.
Someone more familiar with the platform might have an explainer?

EDIT: @toddumptious post above seems to confirm the OG meta mangling? Odd.

[DetectiveGamerTreg]

has anyone put this audio through a spectogram?

[ThatBomberBoi]

yep, it doesnt show anything useful under the spectogram, apart from signs of splicing as we hear in the audio itself.

[toddumptious]

A few of our resident geniuses (geniuii? ) did some good cross examination of the audio in these posts

This is real - Emily Audio

Posting here from the main thread: Spectrogram off of a downloader site, cant see much myself rn but its 3 am here. Im sure @vector_cmdr can do a much better job of that when they’re up haha: [image] Morse code translates to REAL, thus answering the original about page question of Is this real?

This is real - Emily Audio

Chiming in to throw a couple of early observations out before deep-diving: Looks like a standard stereo spectrogram at first glance (with obvious layers and slight screwing/chopping). Strong horizontal energy bands in the middle and clear harmonic structure visible throughout (normal, expected) that can be ‘viewed’ as; The heartbeat (low freq band) is clean and can be isolated more or less perfectly - it’s lovely and anxiety inducing . It just appears to ascend in a normal …

This is real - Emily Audio

Spectrograph in Audio Visualiser. There’s a suggestion of structure in the lower registers, but nothing resembling text. [Emily spectro 2]

This is real - Emily Audio

Chaos game reveals nothing except a faint fractal (except a really nice wallpaper ). So, no sequential data seems likely - though thinking about this more, it’s not particularly evenly structured either. I am putting my finger on the beat/voice scales, but I am happy to be proven wrong. [chaosgame] Edit: CGR comparions: [cgrcomparison]

This is real - Emily Audio

Further notes on L/R channel comparison (visible ~0.79ms lag): Voice band (~160-3000Hz) energy envelopes correlate at .0.92 Heartbeat band (~10-160Hz) energy envelopes correlate at 0.70. The same audio content is in both channels - just artificially decorrelated. After removing the ~79ms lag, raw correlation only rises to 0.28, suggesting a stereo widening plugin (likely Haas effect + phase manipulation?) might have been applied. So the low L/R correlation is likely not a clue either (if ther…

[myrolaw]

I missed so much over the night, wow!

[sheralmyst]

Welcome to the forum @myrolaw

[DevilinPixy]

@vector_cmdr

Edit: CGR comparions:

Am I imagining faintly seeing a possible hexagon, like we had for WT sigils and is commonly used in the website as well? Not sure that helps us with anything though.

[vector_cmdr]

It’s a good question with a probably long and tedious answer - it’s to do with how the chaos game testing is performed.

The shapes come directly from the number of ‘levels’ (vertices) used in the chaos game:

1. Choose N vertices spaced evenly around a circle → these form a regular N-gon.
2. Quantize each audio sample to 1 of N bins (its “vertex index”).
3. Chaos game: start at center, then repeatedly: pick the vertex for the current sample, jump halfway toward it, plot that point.

So the number of levels becomes the number of polygon vertices:

Levels	Shape
3	triangle
4	square
6	hexagon
8	octagon
16	hexadecagon (I had to )

As a function it looks something like:

angles = np.linspace(0, 2*np.pi, levels, endpoint=False)
vertices = np.column_stack([np.cos(angles), np.sin(angles)])

This places levels vertices evenly on a circle - so levels=3 gives a triangle, levels=6 gives a hexagon, levels=8 gives an octagon, etc.
The CGR then maps each quantized sample value to one of these vertices, and the accumulated midpoint positions build up the characteristic fractal pattern inside that polygon.

• An even and mostly patternless fill → values are uniformly distributed across all bins, and transitions between them are random (likely noise).
• Clustering, empty regions, banding → certain amplitude ranges or transitions are preferred, indicating non-random structure (voice, periodic patterns, or deliberate encoding).

Drawing a line between them can be a little hard when you have a sample like ours, where there is voice and heartbeat, but when it correlates with other tests it helps clarify the patterns as likely to only be voice.

Edit: it’s kind of like the sound/vision version of looking at entropy scores for text (etc.) encryption.

[myrolaw]

Thanks! I’ve visited in the past but with this new ARG decided to make an account, i’ll probably continue to lurk as I follow the story though.

[Jupiter.blues]

@myrolaw

i’ll probably continue to lurk as I follow the story though.

Dont worry, i’ve been here 8+ years and I’ve been doing the same. Read everyones posts, hardly post myself. But we’re glad you’re here! Friendliest forum around.

[Turbulentflow]

@sheralmyst, how was this page discovered exactly? Sorry if I missed it but there was a lot going on when the dust was settling.

[ThatBomberBoi]

We found the original page mucking about with the sitemap / WP REST API and added it to the tracker, which picked up this update to the page.

[Turbulentflow]

Thanks, where is the tracker at? Sorry still new to the page but have been tracking the site and doing a deep dive of the ARG since Saturday lol.

[CaptainSteve]

The heart beat sounds in step with the one we hear at the end of every trailer… when the atlas pulses… I think it just means the Atlas is coming alive… or waking up… Meh…

Other than that, listened to the audio a few times, other then echo and reverb, I don’t think there is anything hidden there… Could be wrong though.

[ThatBomberBoi]

Tracker posts to this thread directly:

Skyscraper Haven Map

This website is tracking all the info gathered in the ETARC forums and trying to make it easy for others who are joining the ARG or who want to see what all has been found so far

Would recommended reading through this thread and those linked at the bottom as well:

Project Skyscraper - Welcome / Info

First of all, a warm welcome to our forum and ‘normally’ quiet little place to discuss anything No Man’s Sky. We have received a lot of traffic ever since this likely fan made ARG popped up and literally appeared to ‘kick’ off on here. In hopes to prevent any issues and provide some quick info to get you started, please read up on some basics below: Before you say anything, make sure to bring good manners, we all appreciate those! Have a look around and check …

[Turbulentflow]

What if there is data hidden under the heartbeat in the spectro. Can it be isolated and removed with enough left over to still see it underneath?

[vector_cmdr]

At this point it seems fairly unlikely to me:

(Simple layerings, purely demonstrative):

The heartbeat itself (the ~10–150 Hz band) can be perfectly isolated and extracted and progresses to a near-perfect exponential crescendo.

You can find the extracted heartbeat as well as other (de)layerings, transforms, etc. + different graphing and reports here if you’re curious to see: project_skyscraper_junk/self_introduction_memory_dump at main · vectorcmdr/project_skyscraper_junk · GitHub

Edit:
Of possible note: the track is 16sec and the crescendo heartbeat becomes audible at ~10sec mark.
Two of our hot numbers yet again.

[Loop]

Found it after some digging!

[Mad-Hatter]

Waaaahhh!
Everytime I see an Atlas Pass, I’m reminded of the fact I missed out on one.