It's not an hallucination, it comes from the extended version of the video :
It's "enhanced" and up-resolutioned by AI, so the details are hallucinated - or at least they are highly misleading artifacts of the process. It's less than helpful.
This was a screenshot from a video that came from Maussan Television. I don't think to have the more vibrant colours like in the video The video was entitled "UNAM's Surprising Revelations on the Buga Sphere | The official report" When zooming in on the trees and foliage I see white, silver and jet black coloured elements that I see in Unreal Engine. Happy to be wrong and learn.
Could you please draw some circles and arrows on the image to show us?
You seeing things is not helpful if you can't show what you are seeing.
Here's a version of the video without watermarks
Of Course. I'll have to zoom in as don't have any specialist software is that OK?
Best I can do with what I have. If you want to see the video link I'll put a timrstamp so you can view it with more clarity.
Thank you. Appreciate the help.
I think pareidolia is getting the better of you. The video is compressed and pixelated which makes it easier to read too much into too limited info: Your "repeats throughout" examples don't actually repeat throughout. Personally I see no reason to think its CGI. These would be detailed, high-poly CGI trees, which in my experience of using tree models in blender, would not be using repeated patterns anyway. Its just compression causing pareidolia, including the "owl face" which I don't even see nor do I agree it repeats.
You quoted "Superficially looks like" and "AI hallucination loosely based on the original" which can imply the video is completely AI generated, which it is not. My comment was intended as a precision to avoid misleading people who read yours without the full context.
@Fin , in post number133 (above), shows pictures of the equatorial band tilting from side to side depending on the direction of travel, yet the dark spot stays resolutely pointing at the ground. That indicates that the dark bit is a ground reflection on a closed metal sphere, rather than an open bottom.
I think the ball sometimes appearing to be in front of those blades is just an artifact of the low resolution, motion, and compression. Example below.
if this is not a drone ... xD
If this video would be manipulated in a simple way, then the prnu of the frames with the ball should not be compatible with the prnu of the frames without the ball. So i created prnu file for frames with "ball" and a prnu for without the ball. And the comparison is negative (PCE Match Value of 16 is super low). Maybe i get something wrong here but for me that should be the case, or where is the mistake in my thoughts?
Hi Sure. You can see what has been inserted to give the impression of gaps in the trees But I believe they are elements seen in Unreal and the like for the same purpose.
Also during the whole video, there is ZERO movement of any of the trees or bushes. . The only things that move are the background sky that jitters at times and the "sphere".
It would be a very strange tree indeed that did NOT have gaps in the foliage, wouldn't it? But the thing that seems the most unreal in that picture are the odd "plumes" that stick out over the top of the mass of trees, although I cheerfully admit that I'm not familiar with the foliage that grows at that location. Are these things real?
Tbh it cant be full cgi as the prnu pattern changes drasticly on some moments. I dont know why tho. You can clearly see, that the noise fingerprint changes on some moments. First i thought it happens when the phone is switching the sensor depending on the zoom level, but i think it doesnt fit as on other zooms that does not happen.
it's a different kind of tree.
Plants don't move when there's no wind.
Windless days are also ideal for flying drones.
It would be a very strange tree indeed that did NOT have gaps in the foliage, wouldn't it? But the thing that seems the most unreal in that picture are the odd "plumes" that stick out over the top of the mass of trees, although I cheerfully admit that I'm not familiar with the foliage that grows at that location. Are these things real?
No matter how dense you would still see light coming through the canopy and gaps of varying light around the middle-not a solid block. I've lived next to rainforests and worked in an office over looking one, you don't see a solid block of black just behind the edge
I've also never seen a limb or bush that grows down then decides to grow upward at a 75 degree angle.It would be a very strange tree indeed that did NOT have gaps in the foliage, wouldn't it? But the thing that seems the most unreal in that picture are the odd "plumes" that stick out over the top of the mass of trees, although I cheerfully admit that I'm not familiar with the foliage that grows at that location. Are these things real?
I
I've yet to see anything you've posted that is demonstrably fake. This image you posted with a white line does not seem to indicate anything out of the ordinary on a highly compressed image of foliage. Either clearly demonstrate your theory, or stop it.
Can you explain how you computed the PRNU, and what you used for a reference.
PRNUs are made with Amped Authenticate and i used the averaged PRNU from the whole video against each frame of it (not really each, but like a small group of frames like seen in the screen).
This is a highly compressed video. PRNU analysis really needs video stright from the camera. Doing this on a video downloaded from social media is just going to be compression artifacts.
