'Two things always been off-limits I still say, women and children'
A while back we posted about 6:16 in LA, actually we have multiple times. Back in September we posted about how if you record the audio raw from instagram, export it the right way(616hz sample rate, 24pcm) You can upload it to audiotag.info and the results are that 6:16 in LA is tagged with the songs you see in images 1 through 4. There's other ways to see that there is hidden data in the track that we'll get to in a longer breakdown.
When 6:16 in LA was posted it was at least initially a Dynamic Dash stream. That is the song is uploaded in segments, or packets. When recording raw via audacity, with keeping focus on instagram in a browser(desktop only) Something interesting happens. Different versions of 6:16 in LA were sent out.This required that certain event listener responses gave the right response back. For example, you couldn't tab away, you couldn't click the mouse focus off of the 6:16 page. This get's kinda technical and boring, but the details will be uploaded this week. The fun part is, the song never ends, and when you adjust and export certain parameters you start to see there's data hidden via steganographic methods that are particuarly suited for a dash stream. Above are screenshots of the results.
This was found out, kind of by accident. It started when 2 of us had some weird stuff happen when working with 6:16 in LA. We weren't workign with Youtube versions, we were from the original on insta. One happened when one of the team tried to upload a raw recording from Kendrick Lamar's Insta post and his Daw spazzed out. My daw didn't even recognize the song .wav, and I couldn't upload. Then when I uploaded to Serato DJ, it went crazy, started flickering wouldn't play playing pausing. . .literally bugged it out, this has only happened once, that time. Also, basic screen recording was having issues, that's what led us on this weird path of wondering if something was going on here. I'm going to upload details of some of the files tonight or tomorrow, but it actually looks like potentially being pointed to resources for cryptography and info on Machine Learning(bot stuff). It looks like 6:16 in LA may have been telling us about the BOT problem. This part is still theory because it's all kinda weird frankly. We've been working on this with people outside the team as well. Recently I watched a mini documentary about the Kryptos Cipher at CIA headquarters. It talked about how the creator of the cipher/statue told those that had solved steps 1-3(4 remains unsolved as of today), that they did get the right answer, but not the right way. I used Binwalk to spot that there was something going on here. Then I used Bulk_extractor(Kali default version does not work, had to compile with specific libraries with cmake before it worked properly) to extract the files. Bulk_extractor essentially ignores headers and is not even designed for use on .wav files, but it did work. We think there's more there, but more on that later. This has been a personal pet project of mine, but everytime we try to get people into this, it gets a little crazy, hopefully someone smarter sees this, I'm far from a pro at this stuff. Here is a link to a youtube video we made on this topic(the song tags). The voicechanger is kind of hard to understand so this is going to be redone to be a easier to understand with better editing in the near future, but there still is good info about the why and how this works in the video. The explanations in this video apply to the files as well.