NYT: GIMBAL Video of U.S. Navy Jet Encounter with Unknown Object

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

I take it these are measured using weather balloon soundings so can't be measured continuously. Going from 95 to 120 knots over the course of a few hours doesn't seem that surprising.

It's the huge increase in wind speed from ground level up that surprises me. I guess I missed that phenomenon in Earth Science class.
 
Some observations with annotated video demonstration... All Gimbal-target rotations start simultaneous with an upward bump of the FOV scene, as if the camera was jolted. The target starts to rotate if and only if there is a FOV bump.



Most target rotations correspond immediately with roll-axis rotations of the ATFLIR jet (center icon) such that one target rotation immediately follows a roll-axis rotation and a second target rotation is simultaneous with a roll-axis rotation and both stop simultaneously.

So there are tight correlations between bumps, Gimbal-target rotations and roll-axis rotations of the ATFLIR-mounted jet. All these correlations suggest that Gimbal-target rotations are directly tied to camera events.
 
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.

No, not common at FL200 but certainly not unusual. Very common at FL330 and thousands of airliners fly in much higher winds on a daily basis. Your speed through the air is unaffected by wind, just your speed over the ground. Without instruments or time/distance calculation, in the aircraft, you can't tell there is any wind at all.

Click here to see current global winds at FL340. The site is very interactive, click the hamburger (three lines) bottom left for the menu. 250 hPa is about 34,000 feet, 500 is about 18,000. Blue is calm, white is 250kts

Winds aloft are reported only four times a day because, in fact, the do not change very fast above 18,000 feet where the flight levels begin. Huge air masses just can't change speed or direction very fast (think F=MA).

Can the jammer create false targets on radar?

Yes. Even back in the '70s, we used pseudo-random jamming to put moving targets all over the screen, so SAM operator couldn't tell the real one from noise. "Range gate pull off" was another common technique, which fooled AAA radar into thinking the aircraft was well behind where it really was. Today, code-injection can even be used; that sounds sophisticated, but even self-driving car developers worry about it.

Have you heard of radar reflectors inside balloons, perhaps used as decoys?

Weather balloons carry a radar reflector (it hangs below the balloon with an instrument package) so they can be tracked by radar and winds aloft can be measured.
 
Last edited:
Yes. Even back in the '70s, we used pseudo-random jamming to put moving targets all over the screen, so SAM operator couldn't tell the real one from noise. "Range gate pull off" was another common technique, which fooled AAA radar into thinking the aircraft was well behind where it really was. Today, code-injection can even be used; that sounds sophisticated, but even self-driving car developers worry about it.
Content from External Source
That's very interesting. What about faking apparent velocities on radar ? It's been reported that during the 2014-2015 incidents targets were seen on radar exhibiting hypersonic velocities but I haven't heard of any visual confirmations of those observations. I'm wondering if those radar tracks could have been spoofed by ECM somehow.
 
What about faking apparent velocities on radar?
Sure, common technique. But since we haven't seen the radar recordings can't say for sure. And anyway, are we talking airborne radar and if so Hornet's or Hummer's? Or was it a ship's, and if not why not? If there was really something there we should have all kinds of corroborating evidence. Failing that, to claim it was ECM from a UFO is the least plausible explanation.

Also worth pointing out that there are all kinds of electronic and physical phenomenon that can cause strange radar and IR artifacts. And of course, there are "oops, sorry, didn't mean to do that" such as the time I shut down Las Vegas Approach for a few minutes <blush> during an exercise near NAS Fallon.

Radar fundamentals and ECM overview. (short PDF) "Generally there are angle deception, range deception, and velocity deception jamming against tracking Radar."
 
Last edited:
Absolutely, common technique. But since we haven't seen the radar recordings can't say for sure. And anyway, are we talking airborne radar and if so Hornet's or Hummer's? Or was it a ship's, and if not why not? If there was really something there we should have all kinds of corroborating evidence.

Details have been very sketchy so I'm not sure of the answer to those questions but I think at least some of the observations were made by the new AESA radars in the Hornets.

As for the radar data itself, I've never seen the military release that, it always seems to be classified. So all we have to go on are the statements of witnesses.

Failing that, to claim it was ECM from a UFO is the least plausible explanation.

Why do you say that ? Of course I'm suggesting the "UFO's" were entirely of human origin, either a classified US project testing their tech against these pilots or an adversary such as drones launched from a Russian submarine doing the same.

What do you think is a more likely explanation for seeing unexplained (ie. not like a known missile launch) hypersonic targets on radar around a carrier group ?
 
Why do you say that ?

I meant aliens, but my answer is the same for your hypothetical classified project or Russian (or Chinese?) test: extremely low probability and zero corroborating evidence.

Instead of making up explanations, why not let the evidence show the path to investigate? Lacking evidence the answer has to be "so far we don't know, let's keep looking."

What do you think is a more likely explanation?

Malfunction, misinterpretation, or fake. One of the clips that does sound like real crew chatter seems to me to be a clip that starts just after one guy said to the other, "Weird return on the FLIR, looks like a UFO." And the other crewmember laughs and launches into a fake UFO play-by-play that was intended to give the ready room a good chuckle when they played back the video during the mission debrief.
 
Last edited:
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
Glare rotation in the twin-engine jet is very much distinct from rotation of the jet...



And the dark radial arms that show the rotation in the comparator are just barely visible emanating from the Gimbal target. So, apart from greater degradation of resolution in the Gimbal footage (due to both file copies and distance to the target), there's no big difference between the Gimbal footage and this known case of glare rotation from an ATFLIR camera. So, imo, there is sufficient evidence of glare in the Gimbal footage.

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.
Not all the time. The ellipticity of the pitch-black region in the comparator changes and sometimes is about the same as the Gimbal target. In contrast, the Gimbal target is dramatically further away, most likely tens of miles, and so would show less variation in shape. The twin-engine comparator was very close to the camera.
 
Here's a forever-running GIF of the horizon-marker stabilization analysis...


With annotations, then without, ... Stabilizing to the clouds, as I had done before, was less effective at revealing the perfect relationship between bumps and rotations because in holding the clouds stable, the best marker of the bumps, the clouds bumping up, was suppressed. Here, only horizon marker is stabilized, and is held perfectly stable.

Clearly, all target rotations start with a simultaneous bump-up of the scene, which means with a slight jolt to the camera. So it seems the rotational device is a bit crude, it slightly jolts the camera when initiated. Also, this shows that the big jerk near the end is not exceptional, it's just the largest among several bumps.
 
Here's loops of the start of the main rotation. Things to note:

Gimbal-5314-PP1.gif

  • The horizon indicator does not move.
  • The horizon (the clouds) does a little jump.
  • The sky banding that starts of more of less vertical rotates in sync withe the object (the glare)
  • That rotation is most noticeable in the upper left corner, and is hard to see elsewhere, but it's there.
 
This I hope is a better demonstration of the camera-bump and target-rotation correlation. To reduce visual overload, I've reduced notations from three words to just 'rotate', and it appears only when the target is rotating. Noticing the bumps is facilitated by a static horizontal line...



Keep your eye on the yellow horizontal line to see the upward bumps and seesaw motions of the scene of clouds. Then notice that the 'rotate' notation marking concurrent target rotation appears exactly as the scene bumps up and seesaws, such that each initiation of the four rotations of the target precisely corresponds to a scene bump.
 
Do we understand the purpose of the parallel lines that enclose the target? They change length and move outward, do ?they indicate the status of the target lock? It seems the bumps are either a result of or a cause of lost or less sure lock on the target.
 
A likely illusion that has misled some is that the Gimbal target slows down. I devised what I believe is an effective way to show why this happens (as the LOS approaches 0˚, the flight path transitions to flying directly at the target, during which it should appear to slow down). This modeling is a rough draft, and it gives some insight into the big jerk we see at LOS = 0˚, we can see the orientation mechanism has to reverse direction as LOS passes 0˚.



However, in doing this a problem arises... the clouds in the Gimbal video move from left to right, yet, as this draft video shows, in modeling the screen-data-dictated trajectory we've agreed, it seems the clouds should move in the opposite direction, from right to left. What gives? Rotating the model FOV makes no difference. Am I making a modeling error? Is our deduced trajectory wrong? Or is the target perceptibly moving? In this modeling, I'm assuming, as Mick as concluded, that the target is so far away that its perceptible motion is null, and is further than the cloud bank.
 
Do we understand the purpose of the parallel lines that enclose the target? They change length and move outward, do ?they indicate the status of the target lock? It seems the bumps are either a result of or a cause of lost or less sure lock on the target.

The lines show the tracker gate, indicating target acquisition and tracking. It looks like the bumps throw off the tracker a bit, which is good to know. If the bumps are detected by the IMU, the tracker or the auto-alignment ought to compensate for them.
 
I'd always assumed the clouds were further away, as that's the only way it works.
You're right, you didn't say the clouds were closer. I'd misunderstood. Sorry, my bad! Now it works. :)

That's actually a cool model confusion to go through, as it makes clear how amazingly we can tease out necessary features of the scene from the given information in the footage.
 
Last edited:
Do we understand the purpose of the parallel lines that enclose the target? They change length and move outward, do ?they indicate the status of the target lock? It seems the bumps are either a result of or a cause of lost or less sure lock on the target.
Notice on the first bump-rotation instance, the tracker gates do not adjust. So the bumps aren't caused by changes to the gate (the two lines fitted to the target). Also, lock is only lost, and just for an instance, on the last big bump, before that the quivering of the gate is determining the size of the inner gate (which is the tightest fit on the target).
 
Here's loops of the start of the main rotation. Things to note:

Gimbal-5314-PP1.gif

  • The horizon indicator does not move.
  • The horizon (the clouds) does a little jump.
  • The sky banding that starts of more of less vertical rotates in sync withe the object (the glare)
  • That rotation is most noticeable in the upper left corner, and is hard to see elsewhere, but it's there.

Just repeating this without contrast enhancements, as someone suggested it might introduce "contrast and brightness artifacts" - so rather than argue about what those might be, this demonstrates the banding rotates in the original video.
Gimbal-Original-Lossless-5313-pp.gif
 
Just repeating this without contrast enhancements, as someone suggested it might introduce "contrast and brightness artifacts" - so rather than argue about what those might be, this demonstrates the banding rotates in the original video.
Gimbal-Original-Lossless-5313-pp.gif
Of the four examples (large and small contrast enhanced and large and small original contrast), the banding rotation is most obvious in the small contrast-enhanced version in your quoted content. The smaller version has smoothed/blurred out a lot of distracting noise.
 
Of the four examples (large and small contrast enhanced and large and small original contrast), the banding rotation is most obvious in the small contrast-enhanced version in your quoted content. The smaller version has smoothed/blurred out a lot of distracting noise.
That's somewhat browser dependent. Here's one I've shrunk in Photoshop with the original and contrast enhanced side-by-side


pp1_Lossless-148.gif
 
An interesting fact is that the clouds do not come to a dead stop when LOS = 0˚, whereas they inevitably do in the modeling I'm working on because my clouds are static. So I guess we can attribute the slight amount of motion when LOS = 0˚ to, at least in part, the wind-blown motion of the clouds.

I'm see a big difference, via both Firefox and Chrome, between these three smaller versions....




And the far-superior version (superior in the sense of fleshing out the banding rotation) is the one that the forum software is shrinking in the quote.
 
And the far-superior version (superior in the sense of fleshing out the banding rotation) is the one that the forum software is shrinking in the quote.
That's actually using a different contrast method - a "dehaze" in Photoshop's camera raw, applied twice, I think.
 
Here's a zoom on an updated rough draft, forgot to make visible the top-overlaid screen target and the degree markers, takes so long to render as I'm working with a stack of huge graphics in a 1920x1080 Premiere Pro sequence, nested and then rendered through a 1280x720 sequence, otherwise I can't get it to render correctly. But the right modeling is becoming apparent to explain the slow down...


The speed is 2X the original. The pilot's screen is on the lower right and the real world 'out there' is to the left of it, with the lines of sight from start to finish of the Gimbal footage. A possible comprehension problem of this model is ET believers will say, "Well then the UFO is hovering in place, that's even more extraordinary than slowing down." The answer to that requires elucidating more about very-large distances and the truncation thereof for the convenience of graphical display. One jet trailing another jet 30 miles away sees that jet ahead of it as if it was still.
 
In this intro to an upcoming video, I try to consolidate the skeptical case against the Gimbal UFO, spanning both the particulars of the case and a meta-level epistemic criterion skeptics apply, into one minute...



In as few words as possible, I make the case that skeptics don't have to prove what the object is. A lot of people seem to think, unless we've proven it's a jet, the evidentiary weight of the ET hypothesis remains unscathed. Not so! The case for skepticism of an extraordinary claim need only rest upon demonstrating a lack of corresponding extraordinary evidence. Showing a mismatch between the magnitude of a claim and the magnitude of evidence proffered for that claim is the minimum criterion for 'debunking'. If the evidence could be something ordinary, it cannot be extraordinary. I believe the skeptics have easily satisfied that criterion with respect to the Gimbal UFO.
 
Here's a simple GIF animation running forward and backward on the sudden jolt to the UFO at the end of the Gimbal footage, where the target almost breaks lock. This is stabilized to the horizon marker. Watch the clouds bump up with the target's jolt...



The clouds bump up precisely as the target bumps up and rotates. The initiation of the motion of the clouds and of the target seem to be exactly simultaneous and thus the target's rotation corresponds to a jolt to the camera. And that is true for all four target-rotation intervals, ie, each of the target's four rotations initiate with a bump up of the clouds, as I've shown above. In fact, the target starts to rotate if and only if the clouds simultaneously bump up.
 
With tracers added...




And here I place paths marking the exact courses of the tracers. The bump to the clouds has roughly the same shape as the path of the target, they all experience a sudden push toward the upper right, and then all fall back in the reverses direction, then stability is restored...



The main difference in bump-path shapes is that the clouds bump path is smeared out along its horizontal path across the FOV, whereas the target has to be kept in the center. But it's clear that the motion of the target -- and thus by illusion seems to be motion of the UFO -- is a consequence of a jolt to the camera inducing a global bump to the entire scene.
 
Last edited:
A likely illusion that has misled some is that the Gimbal target slows down.

I was over on Reddit arguing yesterday and someone mentioned the slow-down of the object:

"The white lines on the left and right of the object indicate where the object would have been if it had continued at the speed it was traveling at previously. It slows down as clear as day. The clouds and the tracking markers do not slow down, when the tracking markers adjust you can see the clouds adjust."

To which I asked:

"In this image you're saying the left line is supposed to indicate to the pilot where the object was expected to be based on how fast it was calculated to have been traveling on its past trajectory?"



Is that what the 2 short vertical white lines tracking the object really indicate? I don't recall ever reading that.

I suspect Mick is going to pop in and tell me I need to go back and read the entire thread, but I did a few searches and didn't come up with anything.
 
I was over on Reddit arguing yesterday and someone mentioned the slow-down of the object:

"The white lines on the left and right of the object indicate where the object would have been if it had continued at the speed it was traveling at previously. It slows down as clear as day. The clouds and the tracking markers do not slow down, when the tracking markers adjust you can see the clouds adjust."

To which I asked:

"In this image you're saying the left line is supposed to indicate to the pilot where the object was expected to be based on how fast it was calculated to have been traveling on its past trajectory?"



Is that what the 2 short vertical white lines tracking the object really indicate? I don't recall ever reading that.

I suspect Mick is going to pop in and tell me I need to go back and read the entire thread, but I did a few searches and didn't come up with anything.

The track gate is always in the center, it only changes in size. The closed-loop tracker commands the servo to rotate the camera to keep the target centered.
 
The track gate is always in the center, it only changes in size. The closed-loop tracker commands the servo to rotate the camera to keep the target centered.

"The white lines on the left and right of the object indicate where the object would have been if it had continued at the speed it was traveling at previously."

So, this is an incorrect statement then?

The track gate doesn't function as a visual indicator of where the target is predicted to be, it's an indicator of how much the target is centered in the camera?

Is it even fair to suggest the target slowed down or sped up based on how much it's centered within the track gate?
 
"The white lines on the left and right of the object indicate where the object would have been if it had continued at the speed it was traveling at previously."

So, this is an incorrect statement then?

The track gate doesn't function as a visual indicator of where the target is predicted to be, it's an indicator of how much the target is centered in the camera?

Is it even fair to suggest the target slowed down or sped up based on how much it's centered within the track gate?

If it's tracking well, then the target should stay in the center, but if the camera shakes or the target suddenly speeds up or slows down, then it'll shift off center. As long as it stays in the track gate, the tracker will try to recenter it.
 
If it's tracking well, then the target should stay in the center, but if the camera shakes or the target suddenly speeds up or slows down, then it'll shift off center. As long as it stays in the track gate, the tracker will try to recenter it.



So, it's fair to say igoddard's excellent animation clearly illustrates that it's reasonable to say the loss of tracking is due to camera shake and not the target slowing down?

I think I got it now, thanks for helping me wrap my head around this.
 
So, it's fair to say igoddard's excellent animation clearly illustrates that it's reasonable to say the loss of tracking is due to camera shake and not the target slowing down?

Correct, although I wouldn't call it "loss of tracking" since the tracker doesn't lose lock.
 
Here's loops of the start of the main rotation. Things to note:
Gimbal-5314-PP1.gif

That's somewhat browser dependent. Here's one I've shrunk in Photoshop with the original and contrast enhanced side-by-side

pp1_Lossless-148.gif


The rotating bands seem easier to see taking a larger slice of the footage and speeding it up...



This is sped up 6.5X, with a light 9% contrast boost (which isn't even necessary but helps a bit). Also, giving it a large black frame allows the eyes to better see the subtle contents by removing the glaringly bright forum-background color otherwise immediately surrounding.

Here's an even longer slice of the footage @ 6X speed and de-saturated to eliminate color noise. This rolls through 3 rotational phases, showing that the background 'bands' rotate exactly simultaneous with each target-rotation phase.



Here's horizontally stabilized @ 6X speed. Notice that the three target-bands rotation phases 'snap' into motion... snap... snap... snap... These snaps are simultaneous with the bumps I've highlighted above...



The 'snaps' and simultaneous bumps suggest there is stiction in the rotational mechanism, ie, a slight frictional resistance to initiating the rotation mechanism that slightly jolts the camera once the mechanism pushes through that resistance.
 
Last edited:
A likely illusion that has misled some is that the Gimbal target slows down.



I'm not sure what the intent is behind this video (or honestly, if it's even a valid way to visualize this), but based on the content of the channel my guess is that it's likely being used to support the idea that the object slowed down:



Though, it seems to me that it could instead support your animation\concept.

Thoughts?
 
I'm not sure what the intent is behind this video (or honestly, if it's even a valid way to visualize this), but based on the content of the channel my guess is that it's likely being used to support the idea that the object slowed down:

It shows the apparent motion of the object relative to the clouds. The apparent motion is a function of the actual motion, and the motion of the jet with the camera. The motion of the jet with camera changes from being at an angles with the object to flying towards that object. This accounts for the apparent slow-down.
 
It shows the apparent motion of the object relative to the clouds. The apparent motion is a function of the actual motion, and the motion of the jet with the camera. The motion of the jet with camera changes from being at an angles with the object to flying towards that object. This accounts for the apparent slow-down.

So, it's fair to say this panoramic-stitched video is very misleading about the true nature of the event.

Do you think there is there anything to be gained by this method of analysis or does it just simply look cool?

Recently, I've done the same sort of thing with other a couple of other UFO videos, but where the camera is more or less stationary because it was handheld. Mostly because...it just looks cool.
 
So, it's fair to say this panoramic-stitched video is very misleading about the true nature of the event.

Yes, because it really makes it look like the camera is not moving.

Recently, I've done the same sort of thing with other a couple of other UFO videos, but where the camera is more or less stationary because it was handheld. Mostly because...it just looks cool.
How did you do it?
 
So, it's fair to say this panoramic-stitched video is very misleading about the true nature of the event.

Do you think there is there anything to be gained by this method of analysis or does it just simply look cool?

Recently, I've done the same sort of thing with other a couple of other UFO videos, but where the camera is more or less stationary because it was handheld. Mostly because...it just looks cool.

It makes more sense when the camera is stationary so you get an actual panorama, or if the camera is moving and looking at a flat surface like the ground so you get one large stitched image. Otherwise, if the camera is moving and looking at objects at different distances, you get motion parallax like this
 
How did you do it?

Feel free to delete this post if it's junking up the thread.

The first time I tried this (youtu.be/xTNz4nZq1DU) I used an old copy of Autopano to generate a still image panorama from the source video.

Then in After Effects I stabilized the same source video twice, once stabilizing on the object and then again stabilizing on a still object in the environment.

Then in AE took both those stabilized sequences and layered them over the still panorama image.

For the sequence that was stabilized to the object, I animated the position of the still image panorama to align to the stabilized sequence.

For the sequence that was stabilized to the environment, I animated the position of the sequence to align to the still image panorama.

Playing with UFO videos like this is pretty new to me, so aside from looking neat, I have no idea what my point is.

I mean, stabilizing UFO footage can be useful, but these panoramas I'm not so sure about.

The second one I tried (youtu.be/WtZZ8JWCOlI) just happened to be from the inside of a moving plane, apparently. Same method, but hand stitched the panorama.

Like I said, this is new to me, I'm just trying to sharpen my skills here. So, although the tracking and stabilizing is a little rough at times, it has been getting better.

And yes, I know the first source video is a plane and the second source video is a joke.
 
Getoffthisplanet, here's an updated version of the go-slower-illusion demo you cited...



We can deduce the flight path and the camera's line of sight (LOS) to the UFO from the screen data. This models both the ATFLIR-jet's flight path and how the components of the scene interacted via the camera's LOS to produce the Gimbal footage, simulated in the tilted square box near the UFO.
 
Back
Top