1. Scaramanga

    Scaramanga New Member

    "Twas brillig, and the slithy toves
    Did gyre and gimble in the wabe:
    All mimsy were the borogoves,
    And the mome raths outgrabe...."

    'Jabberwocky'....proof that Lewis Carroll had access to secret Pentagon videos long before TTSA.
  2. Gerard

    Gerard Member


    I just watched your latest video on the Gimbal.

    I understand the notion of glare (ie. internal reflections in the optics) rotating due to rotation of the camera but at this point my question is what glare ? I thought the latest consensus was that the bright halo around the object was an artifact of image filters (specifically an unsharp filter). If that's the case where are we seeing any actual glare in the video ?

    Also your claim that the entire object shape is only determined by glare seems a bit improbable a priori. Do you have any examples of other IR videos where such an effect occurs ?
  3. Mick West

    Mick West Administrator Staff Member

    The black bits are glare. The camera is in BLK mode, meaning black is hot. Black is the "brightest" IR, and is causing a glare. The white "halo" is just the sharpening artifact.

    There are examples in the first post.
  4. Gerard

    Gerard Member

    So you are saying that the image contains nothing but glare.

    To have glare don't you need a primary signal and shouldn't the primary signal normally be stronger than the secondary reflections within the optics that constitute glare ?
  5. Mick West

    Mick West Administrator Staff Member

    Look at the sun.
  6. Gerard

    Gerard Member

    I usually try to avoid that because I value my eyesight but I suspect that if I did the strongest signal would be that of the sun itself and not from some reflection (though reflections can be quite strong they should still only be some fraction of the primary signal intensity).
  7. Gerard

    Gerard Member

    The example in the first post shows a hot spot with structure that is a bit similar to the Gimbal object. I can see this structure being artifacts due to lens imperfections or perhaps diffraction. What I don't see in that example is anything that is clearly glare due to internal reflections nor do I see any rotation effects. There is a different example that does show rotation (the one where we see the outline of the aircraft) but that seems quite different.

    EDIT: OK, there is some glare around the hot spot in the first example but it's not clear that glare is determining the shape of the primary spot itself. There is even less clear evidence of glare in the Gimbal video. (I originally thought the light halo was supposed to be the glare but it seems that was a misunderstanding and without that I'm not seeing much glare.)
    Last edited: Jul 2, 2019
  8. igoddard

    igoddard Active Member

    That range of sizes may be to accommodate different viewing devices, like phones, tablets and desktop.

    Jeremy Corbell seems to claim the Gimbal footage was out before the NYT report...

    See time 7:40, or this link should jump there. That seems to corroborate Jay's claim.
  9. Agent K

    Agent K Active Member

  10. Mick West

    Mick West Administrator Staff Member

    No, it's "Global Security Issues".
    All removed now, and not archived.

    Source: https://twitter.com/Jay09784691/status/1026557498730459136

    Jay found it back in Aug 2018, downloaded a bunch of stuff, including what he says were the original videos. I vaguely remember it at the time, and I didn't download the videos because I though someone said they were just the same as the TTSA web-site videos.

    Back then I did grab an unredacted version of the report from the female pilot (nothing really interesting beyond identifying the pilots)

    Metabunk 2019-07-03 06-00-35.

    So there was certainly a leak. Jay also posted some interesting photos:

    Source: https://twitter.com/Jay09784691/status/1026858994005155845

    DkAizebXsAAFQbG. DkAizy7XsAEaRar.
  11. igoddard

    igoddard Active Member

    One thing we want to keep in mind when assuming signals are "a jet" is there's probably a reason the videos are filed under unmanned aerial systems and balloons. That reason would probably be that none of them are manned aircraft and they might be unusual military drones, aside from the Go Fast balloon.
    • Agree Agree x 1
  12. Mick West

    Mick West Administrator Staff Member

  13. igoddard

    igoddard Active Member

    Check this out, I just noticed camera rotation in the opposite direction of the target rotation concurrent with target rotation, fleshed out in the stabilization I made shortly after the NTY report:

    The target rotates counterclockwise as the camera's FOV rotates clockwise. This underscores that target rotation corresponds with camera behaviors.
    Last edited: Jul 3, 2019
  14. Gerard

    Gerard Member

    Could you please explain exactly how you made this video ?
  15. igoddard

    igoddard Active Member

    Stabilizing to the horizon line indicated by the level of the upper edge of the clouds, going frame-by-frame as needed in Adobe Premiere Pro. So for example, anytime the clouds would bump up, I'd move the position of the upward shifted clouds down in those frames until they align to the level of the same clouds in the prior frames. That removes the motion with respect to the clouds.

    Here's the new copy I just stabilized, this time to the target, just to see what it might reveal...

    Here I created a cutout of the Gimbal-target shape, and placed it fixed in the center, then shifted per frame as needed to align the target in all frames with overlaid mask. So here the target is stabilized. A few frames are imperfect because Premier kept shifting a few frames I already fixed and I don't know why. Not sure if this gives any insight, but it sure feels like the camera is doing the rotation.
    Last edited: Jul 3, 2019
  16. Gerard

    Gerard Member

    Hmm, would it be possible for you to extract the orientation angle of the cloud horizon line and of the object and plot them both as a function of frame # ?
  17. igoddard

    igoddard Active Member

    The clouds are very imprecise and changing on their irregular shapes, so I'm not sure they'd provide reliable data for a mathematical variable. However, perhaps the most reliable metric is the horizon marker in the screen data.

    Here I've stabilized the newly acquired copy to the horizon marker, ie, the horizontal line in the middle of the FOV. Wow, this was insanely simple versus the other stabilizations, requiring only 6 adjustments of position and rotation!

    The clockwise rotation of the FOV while the target rotates counterclockwise is clear to see here. That must surely be the de-rotation mechanism captured in the footage. To believe the TTSA narrative, you have to believe the camera just happened to be doing rotational motion while the target was rotating.
  18. Gerard

    Gerard Member

    What I'd like to determine is the extent to which the object rotation correlates with the background rotation. It's quite hard to tell just looking at the video because you have to watch two things at once. In fact until you pointed it out I hadn't been aware that there was any significant background rotation at all.

    It still seems that the magnitude of the object rotation (which is at least on the order of 90 degrees) is far greater than that of the background rotation.

    I'm not sure how much light the stabilizations shed on this question because they just lock onto one signal so that all the motion appears in the other, but they don't show the relationship between the two.
  19. igoddard

    igoddard Active Member

    Yes, and I just realized, the background, or field-of-view (FOV), rotation corresponds exactly to the rotation of the little icon for the jet in the center of the horizon marker, such that the wings on that jet icon are always parallel to the top edge of the FOV.

    So this background rotation is probably not part of the de-rotation mechanism, it's a rotation that keeps the FOV aligned with the jet as the jet rotates along its roll axis. So the top of the FOV is always aligned with the wings and thus also with the pilot's eyes. And the jet rotates slightly along its roll axis as the line-of-sight (LOS) to target sweeps across 0˚. That makes sense, for as we can see it would be starting to level out of a banking trajectory...

  20. Agent K

    Agent K Active Member

    Right, completely stabilizing the horizon the way you did would make the WSO nauseated because he's sitting in the jet and banking with the jet, so the video needs to agree with his inner ear or he'll throw up. Since the display is banking with the jet too, the stabilized video needs to be rotated the other way to compensate for the display's rotation.
  21. jarlrmai

    jarlrmai Member

    The more we intuit from these videos the more respect I gain for the engineers who design these systems.
    • Like Like x 1
  22. Mick West

    Mick West Administrator Staff Member

    Hmm, I'd always assumed this was understood. I think I might have been too vague in my descriptions before. The little jet icon never moved, and there's a horizon line which is parallel with horizon roughly indicated by the top of the clouds.
    Metabunk 2019-07-04 06-52-51.

    The derotation only keep the horizon horizontal, in that it keeps it horizontal relative to gravity (i.e. keeps it perpendicular to the down direction).

    Since there's potential in confusing the rotation of the horizon with the rotation of the shape, it's important to note there's times when the shape rotates, and the light in the sky rotates, but the horizon and the horizon indicator do not. Most easily noticable is the rotation that happens around 27 seconds, 5312 on the on screen time, when there's a rotation of the glare orientation, the sky light pattern, AND the center position of the glare.
  23. Tom Mellett

    Tom Mellett New Member

    Here is an account written 20 years ago by Tom Mahood which deals directly with how Bob Lazar confabulated those top secret flying saucers to keep him out of jail in 1989.


  24. Gerard

    Gerard Member

    There's obviously a lot of complex stuff going on here but to simplify I think we can say the following:

    1) The object appears to rotate independently of the background.

    There are 4 ways this can happen (that I can think of):

    a) The object itself actually rotates.

    b) The system segments out the target and applies a software rotation to it that is independent of the background.

    c) Mick's theory that the object is actually a secondary effect produced by reflection or diffusion (ie. scattering) within the ATFLIR optics.

    d) The object is actually a defect or foreign matter on or within the optics.

    I think (b) is extremely unlikely because image segmentation is a very complex and potentially error prone operation to insert into a basic visualization pipeline. Also what would be the point in doing this ?

    The problem with (c) is that it requires every part of the image that is seen to move to be a secondary effect, including the core elliptic shape of the object. For that to occur I think you would need some very poor quality or degraded optics.

    Finally (d) can probably be excluded for a number of reasons, one of which is that the object appears to be very hot. Of course I don't know exactly how temperature maps to pixel intensity but I would think that near saturation would correspond to something like a jet engine exhaust and hence far too hot to have on your lens.
    Last edited: Jul 4, 2019
  25. igoddard

    igoddard Active Member

    The jet icon never moved relative to the frame of the FOV, but it does move relative to the stabilized horizontal marker. The issue I was exploring was the clockwise rotation of the FOV as the target rotates counterclockwise, which is fleshed out with horizontal stabilization...

    Explaining that clockwise FOV rotation rested on noticing the jet icon and realizing thereby that horizon-marker stabilization fleshes out the jet's roll-axis rotation, which occurs during the LOS 0˚ sweep over. So a lot of things are happening as LOS sweeps over 0˚, but I don't think this clockwise roll-axis rotation affects the interpretation of the target.
  26. igoddard

    igoddard Active Member

    But Mick has shown nearly identical thermal-signal shapes in FLIR targets of known aircraft,


    and rotation of hot ATFLIR signals (see also center example above),


    without having to appeal to "poor quality or degraded optics."
  27. igoddard

    igoddard Active Member

    This might just be a coincidence, but it's curious to note that both the jet's roll-axis rotation and the target's rotation (during its major phase of rotation) stop at approximately, if not exactly, the same frame...


    Moreover, frame by frame examination suggests that both rotations slow down at roughly the same rate before stopping. However, the target's rotation starts before the jet's roll-axis rotation, and the later jerked segment of the target's rotation does not correspond to any jet-roll rotation. So they're mostly decoupled. But this simultaneous slowing and stopping is curious. I don't know exactly what the make of it, but it may betray some degree of synchronization between the internal camera mechanism we're fleshing out and the jet's roll-axis rotation.
  28. Gerard

    Gerard Member

    The last example you show is just rotating glare. This doesn't explain the Gimbal because there is very little evidence of glare in that video.

    The second example (the back of a twin engine fighter jet) also shows rotating glare but in this case other parts of the image also rotate so I don't really understand what's going on in that case but again it has far more glare than does the Gimbal

    For the first example Mick switches his explanation from glare to diffraction. Note also that the amount of ellipticity is much greater in Gimbal than in that example.

    The reason I'm not convinced diffraction can explain the Gimbal shape is that to do so would require a diffraction limit on the order of at least the smaller half-axis of the ellipse (more likely the larger), which is 6 px. But if the diffraction limit were that large I think the clouds would be less sharply defined. Though I admit that isn't the strongest argument since the clouds aren't all that sharply defined, (I'm not sure I can find a cloud feature that is clearly smaller than 6 px.) It would also mean that in this mode the camera would really not be capable of forming a reasonable image of anything much smaller than full image width, which seems somewhat unlikely given other IR videos we've seen.
  29. Gerard

    Gerard Member

    Incidentally I think an easier way of understanding Mick's rotation effect is that if an optical system has a non-rotationally symmetric point spread function (PSF) whether due to diffraction or geometric aberrations then if you rotate the optics you will also rotate the PSF. If the PSF is large enough to dominate the image then this will cause an apparent rotation of the image.

    I just don't think the later condition is likely to be met in the Gimbal video.
  30. Agent K

    Agent K Active Member

    You got it.

    Why not?
  31. Gerard

    Gerard Member

    Please see my previous post. Basically it seems to me that other image features show too high a resolution to be compatible with a PSF large enough to explain the ellipticity of the target.
  32. Gerard

    Gerard Member

    Additional evidence of this is provided by the gofast video. It seems reaonable to believe that this video was taken with the same ATFLIR equipment used for Gimbal. Also both videos were taken in narrow FOV mode, they should therefore have the same PSF.

    The gofast target is only 2 px high, however assuming that the "Z 2.0" display on the Gimbal HUD indicates a 2x digital zoom this would correspond to 4 px on the Gimbal video. This is still below the minimum 6 px PSF size that would be needed to explain the elliptical shape of the Gimbal object.
  33. Tailspin45

    Tailspin45 New Member

    Very late to the party here, gents, but would like to offer my two cents after an encounter with "one of them" over at Bloomberg who showed his (lack of) expertise by referring to the Gimbal video as radar (sic) evidence.

    I read this thread from the beginning and signed up because I was impressed with the level of the discourse and because I may be able to contribute given that I've been a pilot for over 50 years including eight as Naval Flight Officer assigned to EA-6B electronic warfare aircraft.

    To be clear, my Navy time was in the early '70s, and the Prowler didn't have FLIR. But my wife and I owned and operated a business that offered mock-dogfights and our pilots were almost all Marine F/A-18 pilots, including two Navy Test Pilot School grads and a former XO of TOPGUN. We were an equal opportunity employer, too, so I flew with and became best friends with a gent who had been an Apache test pilot and has lots of FLIR experience. (And I was in VFMA-323's ready room aboard the USS Constellation (CV-64) in the middle of the Pacific when 9/11 happened, but that's another story.)

    So, to cut to the chase, I spent over 1000 hours in a Navy cockpit, and lived and breathed air combat for almost 15 years before we sold the business, which means I think I know what I'm talking about when I say the audio on this Gimbal clip is total BS and obviously overdubbed.

    Casual chat (knowing it is being recorded), after being vectored to investigate something odd, is just not what you would hear on the ICS much less between aircraft. It would be terse and by the book.

    For my money, an internal reflection that over-excited the IR sensor to the point that it created a blob and an "energy field" (slaps forehead), display strobes from the overdriven sensor, a classic demonstration of gimbal lock when the image flips, and the rapid acceleration off the screen when the WSO slewed the pod means we're looking at an aircraft with a FLIR pod that isn't set up right, nothing more, nothing less.

    But, hey, I could be wrong. It could be little green men in a flying saucer. That's certainly what most people seem to think is the most plausible answer.
    Last edited: Jul 5, 2019
    • Informative Informative x 1
  34. Mick West

    Mick West Administrator Staff Member

    Thanks for your perspective!

    It might be that you are thinking of the "Flir1/Nimitz/Tic-Tac" incident when you say "being vectored to investigate something odd." I don't think there's a specific back-story with the Gimbal and Go-Fast videos.

    One suggestion is that Gimbal and Go-Fast might be related to training - firstly in that their labels seem to indicate a type of perceptual problem with interpreting ATFLIR imagery (although that would be more convincing if "Go Fast" was labeled "Parallax".)

    Secondly it might be that the incidents themselves might have occurred during training - specifically training to use the ATFLIR system. Might it be possible that they were simply assigned some time to explore the parameters of (i.e. mess around with) the ATFLIR? So they could just be locking onto random objects, and that might explain the less professional dialog?
  35. Tailspin45

    Tailspin45 New Member

    Yike, you're right. The "go see what it is" event was a different one. Apologies.

    But the unprofessional dialog just doesn’t make sense even in a "go mess around" scenario. The terms they use just aren't part of a fighter crew vocabulary. They'd be talking angles off and over take and lock or lost lock etc.

    The real red-herring for me is the (to paraphrase) 'it's going into the 120-knot wind' bit. How is that even relevant? Maybe, if they were speculating that is was a balloon, but otherwise, you just don't think in terms of winds aloft when everything up there is subject to them and doesn't affect relative motion.
  36. deirdre

    deirdre Moderator Staff Member

    is it possible that basically everything is just less professional these days? My generation dropped the sir and ma'am from of everyday speech. The generation after me dropped the Mrs. and Mr. from most of their speech.
  37. Agent K

    Agent K Active Member

    Maybe speculating about balloons, birds, or drones. Apparently they saw objects flying in a V formation in addition to the Gimbal object.
  38. Gerard

    Gerard Member

    I've wondered about the 120 knot wind thing for a long time.

    Is it common for winds to be that high at 20000 ft ? Because 120 knots sounds like gale or hurricane force to me and you wouldn't think they'd be flying in those conditions.
  39. Agent K

    Agent K Active Member

    I speculated in another thread that the target in the Nimitz Flir1 video was an EA-6B, because it apparently jammed the F-18's radar. Could you look at my posts about that?

    Can the jammer create false targets on radar? Have you heard of radar reflectors inside balloons, perhaps used as decoys?
  40. Agent K

    Agent K Active Member

    This website found an inconsistency between the wind speed and the published date of the video.