DISCLAIMER: This post is potentially VERY NSFW/NSFL!! You've been warned!
Throughout this post I'll reference two videos. They're basically the same video, but one is shorter and higher resolution and the other is longer and has a lower resolution, but both are needed to get the full picture.
VIDEO1: https://www.youtube.com/watch?v=gD_r3D6AU_4
VIDEO2: https://www.youtube.com/watch?v=g6U43kl-knE
VIDEO1 is the main video and I used VIDEO2 as support for some of the things I will show in this post.
Save these videos, because if what I'm about to show you is true, they're getting deleted.
Reddit only allows a limited amount of pictures in a post, so there will be a lot of links to images/videos
Background / motivation
I was never really deeply invested in conspiracy theories (until the whole Epstein thing), but I lurk this sub because it’s entertaining and occasionally raises questions I hadn’t considered.
A few months ago, I came across a post discussing Sandy Hook, which led me to watch several related videos on YouTube.
One of them was Robbie Parker’s interview shortly after the shooting, the one everyone here is familiar with: VIDEO1.
When I first watched it, something felt off to me about the part where he APPEARS to smile before speaking.
I rewatched it several times and I couldn't put my finger on what was bothering me so eventually moved on. But the video stayed in the back of my mind.
Recently I had some free time so I revisited it, expecting that reaction to disappear. It didn’t.
After looping the clip for several minutes and still not being able to pinpoint why it felt strange, I decided to look at it more analytically.
Tooling and approach
I initially searched for existing software that could help with deeper video analysis, but what I found was either too basic or too expensive.
Since I’m a software developer, I decided to write some tools myself which I eventually consolidated into a image/video forensics application. More on that at the end of the post.
Color and brightness inspection
I started with basic color adjustments: brightness, contrast, saturation, and hue. None of these alone revealed anything obvious.
However, one thing became clear: the brightness increases noticeably during the smiling segment.
At first, I assumed this might be a light source or camera exposure adjustment.
What made me pause was that the brightness then drops again as the framing tightens on his face just before he starts speaking.
I’m not an expert in video production, but my understanding is that increasing brightness is sometimes used to mask compression or transition artifacts. That possibility made me curious enough to continue.
Frequency / spectrum analysis
Based on the color behavior, I moved on to frequency-domain analysis.
Running the video through a frequency spectrum analysis algorithm, with some other tools enabled, I got the following results: https://files.catbox.moe/9k224i.mp4
Explanation of the graphs:
- Spectral bands – top left
- Separate histograms for RGB channels – bottom left
- Brightness – bottom right
Notice how the brightness increase, the red peak and green wobble in spectral graph and all the color channels going haywire happens in the same time frame.
For reference here is a spectral graph for an unedited video of a guy dancing: https://imgur.com/QGdvwxz
As I've said, I'm not an expert in video production, so I asked ChatGPT to explain the spectral graph to me. Just in case, I didn't tell it which video it was. Here's what it said:
—————————————————-
Sharp spike in the highest spatial bands (red)
That means a sudden injection of fine detail. Edits often introduce:
- new compression blocks
- new edge statistics
- different sharpening kernels
- different noise profiles
Natural motion almost never hits only the top bands that hard.
Mid-band “wobble” (greens)
That wobble is key. It suggests:
- encoder re-stabilization
- Group of Pictures reset
- bitrate redistribution
- motion compensation settling
In other words: the video codec is “finding its footing” after a discontinuity. You don’t see that with continuous capture.
Low bands (blue) stay calm
This rules out:
- camera shake
- lighting change
- zoom
- exposure shift
An edit often preserves global appearance while changing microstructure. That’s exactly what you have.
Confidence level
From this plot alone: Very strong evidence of an edit. Not a proof in court, but solid technical suspicion
—————————————————-
Face detection / alignment
After doing all of this, I decided to run a face detection algorithm.
The results weren’t especially strong overall, but one detail stood out:
when he's is looking straight into the camera, the detected face aligns well:
But when he’s looking to the side, the detection no longer aligns with his face:
This could easily be explained by:
- limitations of the algorithm
- pose sensitivity
- background similarity
This didn't add anything to the overall conclusion, but it did add another "weird coincidence" into the collection.
Edge detection
Next, I applied several edge detection algorithms and examined the video frame by frame.
In a number of frames, edge detection fails almost entirely on right side of his face.
Again, a possible explanation is that the background tiles are similar in tone to his skin, which would reduce edge contrast, but given the previous results I did not believe that to be the case.
Manual frame-by-frame inspection
At this point, I turned off all of the tools, zoomed in on his face, and inspected the segment manually, frame by frame, multiple times.
WARNING: What follows is a lot of images and short frame-by-frame videos. I will show things that stood out to me the most, but there's too many to show them all.
I started with his face, since that's what the video focuses on, but eventually I started noticing weird stuff on every person in the video.
I'll go through each of them one by one. Most of the pictures will be in regular color, but some of them I'll show with inverted colors. Our eyes and brains are very good at filing in the "missing" details and using inverted colors we can sometimes make unnatural things stand out more.
Let's start with him, since the video wants us to focus on him. First, this frame(VIDEO1 frame 230; VIDEO2 frame 718):
It looks like his head was photoshoped into the picture.
Look at these two frames(VIDEO2 frame 703; VIDEO2 frame 704):
Notice the artifacts next to his mouth in frame 703.
The whole right side of his face is messed up during the whole zoom in segment.
This frame, where his chin is pixel perfect straight, there is some weird artifacts next to his right eyebrow and trails behind his head:
The most messed up picture of all:
Think for a moment what you see in the circle (hint: it's not his hair…)
Two more videos, before we move on to other people. First, a slow motion video of his face. Pay attention to how his face morphs and to his right cheek:
https://files.catbox.moe/3hbk8o.mp4
Finally this video, where a block of his head goes missing for a couple of frames (also, it's not his hair, again…):
https://files.catbox.moe/5p4zn1.mp4
This is it for him, but practically every frame of the video is messed up.
Now to other people in the video
First, the woman that appears in the beginning to "fix the microphones". I think she's wearing a mask or her face has been replaced. Notice the straight lines on her forehead and temples. In the last picture you can see, what i believe to be, a mask strap
Next, the guy in the red shirt and the shorter guy next to him.
Look at the difference between these two frames:
https://files.catbox.moe/f8lfd0.mp4
His hair, his nose, his jaw, his ear and the area around his ear…
Now look at this slow motion, I increased the brightness to make it more visible
https://files.catbox.moe/yj41c3.mp4
Look at how his ear grows, then look at the color of his chin. It's not his beard.
Look at how a "shadow" "spills" next to his mouth, without matching his movement and how his jaw practically disappears when he tilts his head:
Then there's the "preacher" Brett Keller.
These two pictures are 1 frame apart, look at his mouth and chin:
https://files.catbox.moe/7ai8dw.mp4
Also this sequence where it looks like he gets punched in the mouth.
https://files.catbox.moe/sx1bsy.mp4
Look at how his entire face is deformed. When his face swings to the left, notice the artifacts next to his mouth. It might now be very obvious at first, but look at these three frames (VIDEO1, frames 143, 144, 145):
frame 143 – look at the coloring around his hairline, it looks like his face was pasted on and his face looks kinda messed up
frame 144 and frame 145 – in these two frames it looks like his mouth moves like it does when someone gets punched in the jaw
Now we get to the guy moving behind him.
Throughout the entire video, he has no face, except the eyes.
https://files.catbox.moe/dyqaja.mp4
Look at these two frames (it's the same two frames as before for Bretts mouth):
VIDEO1, frame 215 and 216:
https://files.catbox.moe/xv4896.mp4
Same two frames but from VIDEO2, frames 703 and 704:
https://files.catbox.moe/9mnqd8.mp4
Now look at this hand that "wipes his mouth": https://files.catbox.moe/73bx8x.mp4
Why is there, what appears to be, a ring on it, but only for a single frame?
Also, the anatomy of the "hand" makes no sense to me
And speaking of hands:
what is this? https://imgur.com/IsZuJIj
also this: https://files.catbox.moe/vzegfe.mp4
and this: https://files.catbox.moe/pmcfl8.mp4
And the last guy, from VIDEO2, this guy: https://imgur.com/pAO5qlN
Look at his hair, it's short
Here he is, leaving the scene: https://files.catbox.moe/m7x6sf.mp4
Notice how his face gets cut off, but his hair stays. It also looks like his hair has grown.
If i invert the colors: https://files.catbox.moe/lx159i.mp4
It's a bit of a stretch, but it looks like he's wearing one of those white hair nets from clean rooms.
I'll stop here, because this is already pretty long, but there are things to find in EVERY frame of the video. It's absolutely insane!
I'm not going to put any conclusions in your mind, but I think it's pretty obvious what's going on here. Let me just point out THIS HAPPENED MORE THAN 14 YEARS AGO!
Nobody noticed… or they did and were quiet. And I doubt this was the only time it happened.
You can double, triple, quadruple check my findings with the same tool that I used.
About the tool
As mentioned earlier, I ended up building a web app to consolidate the analysis tools I was using.
You can find it here: https://vidana.studio
Everything runs locally in the browser. It doesn’t accept URLs, you need the video files themselves and nothing is uploaded or sent to a server. The files stay on your machine.
For transparency: I use basic analytics to monitor performance and stability, but I don’t collect personal data and I don’t collect or inspect any video content. There are no user accounts and no stored data.
This is an early beta release, and all tools are currently available for free.
I’m mainly sharing this to get feedback on the methodology and implementation, so suggestions, critiques, and bug reports are welcome.
Posted by b-machine
5 Comments
SS: Some time ago I looked at the Robbie Parker video and it didn’t sit right with me. It wasn’t just the smile, there was something more to it. Recently I had some free time and I revisited it. I wanted to do a technical analysis of it, but there were no tools available, so I developed my own custom image/video forensics tool and analyzed the video with it. Here are the results and they’re beyond messed up.
Just one question:
Are you old enough to remember watching him speak live? Because I am; I was actually pregnant with my first child at the time and I remember watching it on TV and finding his affect incredibly eerie and strange – similar to what we have been seeing recently from Erika Kirk.
However, that doesn’t mean that he wasn’t a real person speaking to the camera.
Gently, I want to suggest that this post may be way more telling of your current state of mind, than anything about what happened at Sandy Hook.
Edited to add that, furthermore, it would not make any sense to edit the footage in order to try to make him look LESS emotionally congruent, not more.
Are you good bro?
Have you performed the same analysis on a control video to see what you end up finding? My suggestion would be to apply a healthy amount of objective skepticism to your own work to see how it stands up to data rather than subjective interpretation. I say this as someone who believes what you’re saying.. make it as airtight as possible so you head off the detractors up-front.
I know all too well what it’s like to lose loved ones. Fortunately, in my case, never to violence and not my own child, but when these things are fresh on your mind, you might occasionally smile or laugh, but it immediately goes away once your mind returns to reality. Just the way his smile lingers until he “gets into character” is all wrong. He looks too well-rested. He’s not in shock either, like some have said. Hell, it was harder for me to talk about my dog dying than it is for him to talk about the killing of his child.