GImbal Glare, Rotation, Clouds, and Angles

[TheCholla]

In a small room? This is outside.

You're as always very sure of yourself, but in the end two examples of pod roll in FLIR videos suggest the opposite of what you're saying.
Or do you suggest pod rolls CW in these two examples? One is clearly a plane being seen from above and tracked left to right, like Gimbal.

 

[logicbear]

In a small room? This is outside.

Right, but the point is that the ground and the walls appear to be relatively nearby, and viewed from a fixed point of view. Not the case in Gimbal. In Gimbal you have patterns in the sky, not on nearby reflective surfaces. You can't even argue that the atmosphere is reflecting the object's IR because the field of view is moving. Any unevenness of the atmosphere's reflectivity would not stay constant until the object starts rotating.

two examples of pod roll in FLIR videos suggest the opposite of what you're saying.
Or do you suggest pod rolls CW in these two examples? One is clearly a plane being seen from above and tracked left to right, like Gimbal.

You made the point that it might be caused by the sun in the case of FLIR1. Well it's night time during Gimbal. I don't know if the example in FLIR is caused by exactly the same phenomena as in Gimbal. It doesn't look quite the same to me, regardless of which way it rotates. I think they both look like optical phenomena, not the kind of thing that would rotate with a small target's axis of elongation, if it were a real object, but if one of them rotates a certain way it doesn't necessarily mean the other one has to rotate the same way.

 

[TheCholla]

Right so in other words the evidence is much weaker than advertised, no example for it, just some kind of reflections that do not rotate like they would during the day. Reflections that come from we-don't-know-where since the only source of bright radiation reaching the pod is the object.

I'm not trying to convince anyone here, just testing the waters to see how solid the arguments are for the "glare from a hot engine" theory. And also to hear good arguments against alternative hypotheses, rather than the usual brawl on X.

Before helping with drafting a more detailed demonstration of the alternative theory (close object). We got answers and a lot of insights, despite reluctance to engage.

 

[JMartJr]

Anyway. Light patterns have directly been compared to FLIR1 to prove that rotating light patterns in Gimbal are from pod roll.

In many ways that does not even matter. The fact that the "target" in Gimbal appears to rotate every time the pod would rotate, the fact that it never rotates at an other time, and the fact that every rotation of the target is preceded by a "bump" in the camera show clearly that the rotation effect is caused at the camera/plane end of the line, and not out at the UAP end.

Figuring out exactly what causes the glare to look like it does is a worthy goal, but may never be possible given that we are dealing with a camera/tracking system with features that are classified. But it is not necessary to do that to show that, however the form and shape of the glare is caused, the rotation is a feature of the camera and imaging system on the plane, and is not a feature of the target. As the target does not do anything else mysterious beyond appearing to rotate (but only whenever the camera does), there is no reason to believe there is any mystery here beyond the technical one of understanding exactly how the classified hardware/software of the imaging system caused the shape of the glare. Even if we NEVER fully understand that, we DO know that the video does not show the target doing anything mysterious nor even particularly interesting.

(Note that I am using "glare" here somewhat generically, those with deeper camera-knowing may come to prefer a different technical term. That's fine, whatever term is best is fine with me, and is still a feature of the camera system, not of the target.)

 

[TheCholla]

That's circular reasoning, because a large part of the conclusion that the camera rotates when the object does, has come from the rotating light patterns. Then the pod sim was built to try to fit the roll to the object's angle, using the assumption of a step-rotating pod well beyond singularity where steps usually happen. The glare theory is a circular list of arguments (the "4 observables") that are each supposed to justify weaknesses in the others. I call this is a house of cards.

The bumps are a whole other story, there are alternative hypotheses than the camera "bumping", it's not a regular camera but a fine-tracking system with sensitive internal components. If the object being tracked changes attitude, we can expect the system to readjust.

 

[JMartJr]

The bumps are a whole other story, there are alternative hypotheses than the camera "bumping",

Really? Can you share a couple? Remember that it needs to account for the bump happening every time the rotation is ABOUT TO happen, and at no other time. If it is caused by the UAP/target, an alternative hypothesis has to account for the target being able to reach out and bump the camera.

it's not a regular camera but a fine-tracking system with sensitive internal components. If the object being tracked changes attitude, we can expect the system to readjust.

Can we? If the object stays in the exact same place relative to the camera but rotates a few degrees, why would the camera system need to readjust it's aim?

 

[TheCholla]

the fact that it never rotates at an other time

See this is not a fact, unless we are loose with facts. Even logicbear's objective measurements find it (see the other active thread on Gimbal for more details).

If it is caused by the UAP/target, an alternative hypothesis has to account for the target being able to reach out and bump the camera.

It's not about the target reaching the camera to make it bump, it's the fine-tracking inside the pod that can readjust and cause a bump when the object being tracked changes attitude in the image (like an object abruptly tilting/decelerating in the case of Gimbal). A WSO suggested so back then (something along the lines of "it looks like the ATFLIR reacts to the object changing attitude"). I don't have it now but @Zaine M. may, there is a FLIR video out here of a fighter jet being filmed shooting a missile, which makes the video bump. @Zaine M. I'm sure you know which one I'm talking about.

 

[logicbear]

we-don't-know-where since the only source of bright radiation reaching the pod is the object.

Again, I provided two other possible sources of radiation. Simply the background, if it is unevenly attenuated/scattered by the optics. Or some internal components which might not need to be particularly hot at all because they are nearby, the background is very cold, the sensor is extremely sensitive, and the gain might be cranked all the way up. And again, you have not shown any remotely plausible way for a high contrast object to cause the rotating background patterns, except ones that would rotate with the camera instead, given the fact that the target is small and its position doesn't change much relative to the scale of the image. You have zero examples of what you suggest, and it doesn't make a lick of sense in any shape or form. We should look for better examples as well, but at least there are plausible ways the patterns could've rotated with the camera.

Even logicbear's objective measurements find it

Along with an initial change in shape that can bias the measurements, there's a couple degrees of change over the first 20 seconds that might fit the current model, yes. But more importantly, you see in that image that after the initial change in shape there's not even a hint of the glare angle changing in tandem with the bank angle in the data, or the clouds which rotate even more than the bank angle. Even just that alone shows it cannot be a real object. A real object would have to always rotate immediately in line with the ownship's bank angle changes to cancel them out, as if it were pretending to be glare.

 

[TheCholla]

Along with an initial change in shape that can bias the measurements

You keep saying this but there is no more change in shape at the beginning than later. It changes shape all along as if we see more of it.

there's not even a hint of the glare angle changing with the bank angle in the data

There is also not a hint of your hypothesized dero for pilot's comfort. Which leaves cloud motion angle unexplained, and this is very important for flight paths.

 

[Mick West]

You keep saying this but there is no more change in shape at the beginning than later. It changes shape all along as if we see more of it.

The initial shape changes are asymmetric.

 

[TheCholla]

So now you're saying there is no CW rotation in the first 20sec of the footage.
Sounds like the refinement to the sim was just a waste of time then.

We'll propose something to explain cloud motion angle, that demonstrates that the object was indeed where it was and not a random plane flying away at 19,000 ft, just along the path that makes the clouds look good in the shot. Too bad you have never answered how you estimated with certainty that what we see is exactly the cloud-sky line, which is at the base of the other house of cards, the one about the distant plane.

 

[TheCholla]

it doesn't make a lick of sense in any shape or form.

You seem stumped by the idea that a high-contrast object may create rotating artifacts in the image. Yet there is a big one right there, that rotates with it (the "glowing aura"). It doesn't have reflective surfaces either, but it's there.

So as usual I don't understand how you can be 100% affirmative, without other comparable footage to verify.

 

[Mick West]

You seem stumped by the idea that a high-contrast object may create rotating artifacts in the image. Yet there is a big one right there, that rotates with it (the "glowing aura"). It doesn't have reflective surfaces either, but it's there.

That's a local image optimization done after the sensor. It's nothing at all like the full-frame rotating areas we see in gimbal.

 

[TheCholla]

That's a local image optimization done after the sensor. It's nothing at all like the full-frame rotating areas we see in gimbal.

My bad, I forgot that was post-sensor.

What's your explanation for the light patterns? A few weeks ago it was the non-uniformities on the sensor, still the case?

 

[JMartJr]

The initial shape changes are asymmetric.

Hi @Mick West . That just went well over my head, probably everybody heard the whooshing sound distninctly! If you would expound on what that means and why it matters, I'd be interested.

 

[Mick West]

What's your explanation for the light patterns? A few weeks ago it was the non-uniformities on the sensor, still the case?

The Gimbal ones? Possibly NUC. I don't feel like it really matters

 

[Mick West]

Hi @Mick West . That just went well over my head, probably everybody heard the whooshing sound distninctly! If you would expound on what that means and why it matters, I'd be interested.

With Gimbal the glare starts out asymetric, almost eggplant-like

Then becomes more symmetric, along both axes. This could be interpreted as rotation.

Really looking at these images, an optical artifact (glare) is the only thing that makes sense. You can see diffraction spikes, and considerable changes in the outlines.

It is the weight of the evidence that leads me to have no sense of urgency about addressing the initial rotation quibbles, or this aspect, again.

https://www.metabunk.org/sitrec/?sitch=video&video=Gimbal

 

[TheCholla]

I don't feel like it really matters

Nothing does, but they also do matter a little at the same time.

Initial rotation being eggplant morphing, you should eventually remove that forced derotation in your sim. You should see that only a slight decrease in elevation angle can explain cloud motion angle.

Then it's fun to revisit the rotating glare along the flight path that the aircrew reported.

Nobody is asking you to do anything by the way, I'm just sharing what I see.

 

[Mick West]

Nobody is asking you to do anything by the way

I'm pretty sure someone is asking me to do something

 

[TheCholla]

I'm pretty sure someone is asking me to do something

"you should" is a suggestion here
About what I think would be more accurate given the evidence. Of course you do what you want with your sim.

 

[JMartJr]

it's the fine-tracking inside the pod that can readjust and cause a bump when the object being tracked changes attitude in the image

Can't possibly be, the bump happens before the rotation. This would be consistent with something like the motor(s) kicking on and taking up load to get the pod rotating. It cannot be a response to the target bring about to start rotating.

Edit: hit "save" before finishing my point, so have now finished it!

 

[TheCholla]

Can't possibly be

I respect your position but I'm wary of definitive takes when dealing with a complex system we don't know in details, and for which we have limited footage.

This would be consistent with something like the motor(s) kicking on and taking up load to get the pod rotating.

Would we expect random bumps, or some patterns in the bumps? Because they always go the same way (upper-left jump in the object, slight CW tilt in the clouds). We've done deep dives into this frame-by-frame, it's not so easy to interpret imo.

Simply the background, if it is unevenly attenuated/scattered by the optics. Or some internal components

I wonder if this (exaggerated) modeling of the light patterns could make the big picture clearer.


Source: https://x.com/Oswald1160/status/1770346626843836543/photo/1


Just some observations, that I cannot interpret :

- it's more than a simple rotation around the pod axis, the big reflection leaves the frame during the slow roll, some others fade in fade out.

- internal components: the whole optics and the sensor roll in concert with the pod head. Dero derotates and makes the patterns rotate versus background. Internal mirrors should be centered during the long slow roll, barely moving, and pod ball pitch is changing slightly.

- less pronounced patterns when lock is lost (not shown in this short video), not sure if that means anything.

EDIT: that X post was a response to me noting the the object is the only source of radiation in the image. I meant hot sources in the image but yes there are also the clouds and sky.

 

[JMartJr]

I respect your position but I'm wary of definitive takes when dealing with a complex system we don't know in details, and for which we have limited footage.

It seems to me pretty definite that, if the bump of the camera happens before the rotation, it cannot be caused by a reaction to the rotation, but might be part of the sequence of events that causes the rotation. Are we together that far?

 

[Mick West]

"you should" is a suggestion here

I was referring to Marik.

 

[TheCholla]

It seems to me pretty definite that, if the bump of the camera happens before the rotation, it cannot be caused by a reaction to the rotation, but might be part of the sequence of events that causes the rotation. Are we together that far?

The object seems to "jump" before each bump, so we proposed it may be suddenly decelerating while tilting (it needs to decelerate along the close flight path). And fine-tracking reacts to that.

Do you have an idea why the motors loading before roll would bump the image in the same direction every time? Where are the motors located?

 

[JMartJr]

Do you have an idea why the motors loading before roll would bump the image in the same direction every time?

Perhaps because they are rotating the pod the same way each time.

Where are the motors located?

No idea, but they are on there somewhere! If it is not the motors bumping as it engages, which i mentioned as an example, it is something else and we may never know exactly what because stuff is classified.. But whatever it is, the bump happens before rotation so it is not a reaction to trying to track a rotation that has happened.

The object seems to "jump" before each bump,

I am not seeing that. But I'll go back and look again. in the meantime if you can figure out how to point it out clearly, that would be helpful.

 

[TheCholla]

Here is one. They all look a bit like that, the object jumps up a bit in the lock bars, and the clouds tilt slightly clockwise, at the very beginning (1st frame).
Sharing because you ask, not to start an argument about what causes them.

EDIT to clarify that this is the very first frame of the bump

 

[TheCholla]

For completeness, I found that one in my files, it's the last one. A couple frames I think.
This one the object bumps more to the upper left, the clouds still do the slight CW tilt.

 

[jackfrostvc]

What convinced me other than the synchronisation of the rotation, was Dave Falch's video he then took down which showed a FLIR video of an FA/18 I think, that was flying away. The footage looked earily similar shape to the Gimbal. So much so, people commented just that

 

[Harabeck]

What convinced me other than the synchronisation of the rotation, was Dave Falch's video he then took down which showed a FLIR video of an FA/18 I think, that was flying away. The footage looked earily similar shape to the Gimbal. So much so, people commented just that

Is that footage available from another source?

 

[Mick West]

Is that footage available from another source?

 

[logicbear]

it's the fine-tracking inside the pod that can readjust and cause a bump

Fine tracking alone cannot reproduce the horizon rotation that happens during the bumps. A while back I calculated how much it would rotate the image if a fine steering mirror deflected by a couple of pixels, and it's nowhere near enough. So you need pod roll and/or the dero to rotate a couple degrees, but why would they do so, by that much ? If there's a significant amount of acceleration/deceleration during the bumps then that should show up in my motion data here, but there's no trace of it. If it's only a minuscule/temporary change, then the horizon rotation doesn't make much sense.

there is a FLIR video out here of a fighter jet being filmed shooting a missile, which makes the video bump. @Zaine M. I'm sure you know which one I'm talking about.

Let's see it. I vaguely recall seeing something like that, and I think the horizon didn't rotate.

 

[TheCholla]

With Gimbal the glare starts out asymetric, almost eggplant-like

Then becomes more symmetric, along both axes. This could be interpreted as rotation.

I just tried to isolate the rotation from the top of the eggplant (stem?) from the bottom (the bulge?).

The stem shows CW rotation (with bank):

The growing bottom makes the lower half look like it rotates CCW:

That has caused endless debates about rotation/not rotation. I also wonder how an algorithm would make any meaningful measurement of the object's angle, given it's literally made of two parts going the opposite way.

I honestly don't know which side benefits from this.

If glare fixed in the frame, it weirdly rotates with bank at the top, while growing at the bottom. Never seen one like this in any of the examples.

If diffuse object not fixed in the frame (glare-like but not rotating with the camera), the main part we see from the beginning rotates, but the growing bottom completely changes the appearance of it. Change in perspective is expected from closing on the close-path object and seeing it a bit more from the side, so maybe we see more and more of the front part?

Edit: fixed wording

 

[jackfrostvc]

@Harabeck
There is also this Dave Falch video which shows what seems like a FA/18 looking like the Gimbal video in FLIR mode., especially towards the end

It's cued up to the point I like


Source: https://youtu.be/mk3fG_lREEI?si=SECLGzHyPRXmgOta&t=215

 

[Mendel]

Here is one. They all look a bit like that, the object jumps up a bit in the lock bars, and the clouds tilt slightly clockwise, at the very beginning (1st frame).
Sharing because you ask, not to start an argument about what causes them.

View attachment 87523

EDIT to clarify that this is the very first frame of the bump

so the object does not jump relative to the background
which is why it's a camera bump

 

[TheCholla]

@Mendel fine-tracking causing bumps in reaction to a change in attitude of the object, is a camera bump. Just not from the same cause.

 

[TheCholla]

Speaking of jump, @Mick West 's stitching from the new OpenCV tool in Sitrec retrieves the slight rise in altitude of the object, versus the clouds, during the final rotation.

Observation we had made and included in our paper 3 years ago.

Mick's stitch :

This has already been shown by @Zaine M. in his stitch a few weeks ago. Three independent analyses retrieve it.

Even before debating if the initial slant in the stitch is from a decrease in elevation angle of the camera, or mismatched horizons from the dero, this is incompatible with the distant plane hypothesis.

 

[Mick West]

Even before debating if the initial slant in the stitch is from a decrease in elevation angle of the camera, or mismatched horizons from the dero, this is incompatible with the distant plane hypothesis.

Why? We get the same change in horizon in Sitrec. We know that those angle change exist, these panoramas just demonstrate it differently.

 

[TheCholla]

The object slightly rising above the clouds at the end is incompatible with a distant plane. Unless the random distant plane suddenly gained altitude right before/during the rotation.

 

[Mick West]

The object slightly rising above the clouds at the end is incompatible with a distant plane. Unless the random distant plane suddenly gained altitude right before/during the rotation.

I don't think you've demonstrated that it rises above the clouds.