After the plane disappear there's non movement anymore so compression algorithms could have a easier task and give a sharper image.An observation in the webarchive source video. From the 'teleportation zap' frame onwards, the entire scene is significantly sharper than before the zap. No change of sharpness occurs from frame to frame for the entirety of the video until the zap. Make of that what you will.
That’s not what you’re looking for, though — the drifting cursor is moving with subpixel precision, not snapping from one position in the pixel grid to another. There’s no device or operating system which does this.
this means that the plane footage is one video and the cursor a legit screen recording?
does this mean that the reddit claims of it being a virtual machine or custom os is meaningless?
Aside from the parallax issue, do the deltas in the coordinates imply a distance travelled by the aircraft of only 3.2km in sixty seconds? Because that is only 192km/h, which is extremely slow and possibly below the stall speed for a fully laden 777.Can anyone please explain how video from a satellite of a plane flying within the frame shows ZERO parallax? Is it even possible?
A satellite orbiting between 200-1000 km traveling @ 7.5-10 km/sec. observing an airplane in flight for 60 seconds and panning an area on the ground covering roughly 3.2 km according to the satellite coordinates from the beginning to end of video. Assuming the coordinates are at the center of the video on earth.
Reddit has got to the point of constructing an elaborate system of coincidences and unfalsifiable hypotheses to explain away every flaw in the videos.
In principle it could be a slowed down video, but the aircraft movement looks rather natural speed-wise against the clouds from purely visual judgementAside from the parallax issue, do the deltas in the coordinates imply a distance travelled by the aircraft of only 3.2km in sixty seconds? Because that is only 192km/h, which is extremely slow and possibly below the stall speed for a fully laden 777.
Normal cruising speed is Mach 0.85 or ~900km/h according to several sources scraped by Google.
I am making the assumption that the coordinates are a point on earth that is center in the image. I am simply curious about the lack of parallax. Until this is explained nothing makes me believe this is actual video of a jet. Even if it were from anoter plane close by we should see parallax.Aside from the parallax issue, do the deltas in the coordinates imply a distance travelled by the aircraft of only 3.2km in sixty seconds? Because that is only 192km/h, which is extremely slow and possibly below the stall speed for a fully laden 777.
Normal cruising speed is Mach 0.85 or ~900km/h according to several sources scraped by Google.
maybe im reading it wrong but i cant find where it confirms that satellites have parallax correction software build in?Satellites have software that corrects parallax error. But do we see it in this video it's unclear. We need a real military satellites expert lol.
https://cimss.ssec.wisc.edu/satellite-blog/archives/217
You can't fly a standard rate 2-minute turn (like the aircraft in the video) at that speed; they'd fly a half-rate 4 minute turn instead, or simply choose a convenient bank angle if not under ATC supervision.
It doesn't?
Article: SBIRS is to use more sophisticated infrared technologies than the DSP to enhance the detection of strategic and theater ballistic missile launches and the performance of the missile-tracking function.
SBIRS High (also now simply referred to as "SBIRS") is to consist of four dedicated satellites operating in geosynchronous Earth orbit, and sensors on two host satellites operating in a highly elliptical orbit. SBIRS High will replace the Defense Support Program (DSP) satellites and is intended primarily to provide enhanced strategic and theater ballistic missile warning capabilities. SBIRS High GEO 1 was launched on 7 May 2011.[12] Two SBIRS sensors hosted on two classified satellites in highly elliptical orbit have already been launched,[13] probably as part of the NROL-22 (USA 184) and NROL-28 (USA 200) launches in 2006 and 2008.[14][15] USA 184 and USA 200 are believed by analysts to be ELINT satellites in the family of JUMPSEAT and TRUMPET; TRUMPET has been reported to have carried an infrared sensor called HERITAGE.
I'd love to hear their take on it. The satellite video passes the tests that they usually pick up on (motion blur, consistent lighting, black levels). It's also worth noting that their expertise is not in satellite imagery, and there may be other things unique to satellite imagery that they would misinterpret as fake or not pick up on as fake.
The blog post mentions that when GOES-14 is corrected it uses "infrared imagery to estimate the height of the cloud". There are a few other ways to estimate depth that I mentioned above (in-track stereo or SAR interferometry). This correction doesn't happen on the satellite, but in-software on the ground. The visible imagery is combined with a depth estimate to create a displaced image.
The NRO does make videos, and is always seeking proposals for improving these capabilities.
I've looked more closely at a few versions of this video, and I think the only useful ones are the RegicideAnon video and the Jose Matos video. The Vimeo one is converted from 24fps to 29.97fps, and is missing frames as you mention. The Jose Matos video is exactly the same length as the RegicideAnon video. I suspect that the "original" or "source" video may have been two separate videos that were combined by RegicideAnon into a single side-by-side video (with some vertical black bars/cropping, for some unknown reason). And that Jose Matos just uploaded one of the two videos.
During the initial test of STSS program in 2010, the Missile Defense Agency used two satellites, STSS-1 and STSS-2 to test stereoscopic missile tracking.
Your theory is plausible, but the fact that you can perfectly align the left and right view after a small un-shear of the image means that there is no depth data.
Yes it is? Just paste the overlay to both videos with a small offset?
I meant that if you have a video with the overlay already burnt in it's hard to fake the second image because you have to offset the overlay and restore the background behind it. I wasn't precise enough.
I think we agree on this point but I'm not sure. The cursor and text are also affected by the depth map. Here is a video I made showing the text being affected. The cursor disparity is not fixed throughout the video, but modulates slightly. But this needs a more thorough investigation, it could just be compression artifacts.
The video on the right is not an exact copy of the one on the left, or simply warped by a slight shear, but distorted by a depth map. This could have been handcrafted, to some extent, by using a mix of a gradient plus the brightness of the clouds. Or it could have come from some kind of satellite imaging device that estimates depth. Here is a reconstruction of the depth map using StereoSGBM.
Definitely not for missile tracking, but in the context of mapping and visible light imagery this definitely happens. Here's a demo video showing a product that supports this application. From a UX perspective, I can see how a slight 3D effect would help a user quickly distinguish whitecaps on the water from clouds and planes.
Like what?
1. In-track stereo: unlikely, because we should expect to see more parallax motion from frame-to-frame.
Sorry I don't really get the whole depth map discussion - could anyone quickly elaborate what it means and what we would expect/wouldn't expect, if anyone would be kind enough and has time on their hands?
The StereoSGBM depth map doesn't seem to match the general shape of the clouds.
If you have 2 images of a scene with only the camera location changing between the 2 images you can compute a depth map of the scene. This is an image where the intensity of a pixel correspond to the distance of the object to the camera. That's called stereoscopic 3D.
Thanks very much, got it.If you have 2 images of a scene with only the camera location changing between the 2 images you can compute a depth map of the scene. This is an image where the intensity of a pixel correspond to the distance of the object to the camera. That's called stereoscopic 3D.
For the scene we have we should get something like this :
- the ground is far so it's black
- low altitude cumulus clouds are dark grey because they are closer to the satellite
- cirrus clouds are high altitude clouds so they are even closer and a lighter grey
- the plane might be somewhere between the clouds layers so it's an intermediate shade
