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

Make the clouds as dark as they appear in the Gimbal video. There's a very low contrast between the tone of the clouds and the sky, but your example has a very high contrast.

Nice.

blurred_clouds_w_unsharp_mask_02.gif

I should probably stick something in there to represent the object, but in this case it would be difficult to see the results of the unsharp mask process on the clouds.
 
So, why is Fravor out there saying there is a "force field" around the Gimbal object?

Again, shouldn't he be familiar with this or shouldn't someone at least have told him the aura is the result of image sharpening?
 
If the aura was caused by an unsharp mask process wouldn't/shouldn't Fravor be well familiar with its characteristics or would this sort of minutia be strictly under the purview of the WSO?

The WSO would know this for sure, and I'd expect the pilot to be familiar with it since it's the pilots and/or WSO's who request more image sharpening. How do single-seat F-18E pilots operate without a WSO? Are they trained more than the F-18F tandem-seat pilots?
 
Air speed is 241 Knots, 277mph, so in 10 seconds the jet would have travelled 0.77 miles.
Why is that conversation of knots to mph the conversion we should use?

That's the conversion of CAS to TAS-mph at sea level...


@ http://www.hochwarth.com/misc/AviationCalculator.html#CASMachTASEAS

But that does not match the ATFLIR's altitude of 25k ft, it also does not match the Mach number on the ATFLIR screen of 0.58. But if we enter an altitude of 25k ft, it matches the screen data but TAS-mph goes up to 403 mph.

Sorting this out is necessary to determine the circumference of the circle the Navy jet travels as well as the distance traveled. How can we model the situation correctly if we're using seal-level data for a jet at 25k feet?
 
Why is that conversation of knots to mph the conversion we should use?
241 knots IS 277 mph, regardless of altitude. There's not question there. It's like converting from mph to km/h At the time I wrote that (Dec 2017) I was unaware that the HUD numbers were CAS not TAS (I only realized that when looking at the Go Fast video later), so I was not converting from CAS to TAS, I was converting from knots to mph. The result is incorrect because I should ALSO have converted from CAS to TAS.

The calculator you link to is doing TWO conversions. It's converting CAS to TAS, and it's converting knots to mph. You can make it just do the TAS -> CAS conversion. Like here I'm just doing it in knots.

Metabunk 2019-10-07 10-52-54.jpg
TAS (True airspeed) is what should be used, regardless of if it's in Knots, mph, fps, kph, or whatever. I'm not arguing with that at all. The problem here is when you say:

Why is that conversation of knots to mph the conversion we should use?

You seem to think that converting from knots to mph is some kind of conversion that changes with altitude. It's not. It's just multiplying by 1.15078.
 
TAS (True airspeed) is what should be used, regardless of if it's in Knots, mph, fps, kph, or whatever. I'm not arguing with that at all. The problem here is when you say:

So the correct TAS-mph to use in the circumference equation is 403 mph, not 277 mph?

You seem to think that converting from knots to mph is some kind of conversion that changes with altitude. It's not. It's just multiplying by 1.15078.
No, all I've thought is what I've shown, that CAS --> TAS outputs differ by altitude, and that's true whether it's CAS(knots) --> TAS(mph) or CAS(knots) --> TAS(knots).
 
So the correct TAS-mph to use in the circumference equation is 403 mph, not 277 mph?
403 mph is the correct airspeed to use.

No, all I've thought is what I've shown, that CAS --> TAS outputs differ by altitude, and that's true whether it's CAS(knots) --> TAS(mph) or CAS(knots) --> TAS(knots).

Then there's no problem, it just wasn't really very clear when you said:
There seems to be an error in your calculations based on your assuming 241 Knots = 277 mph, which, however, is only true at sea level. At 25,000 ft altitude, 241 Knots = 403 mph (see my last reply above).
Which looked like you were only discussing Knots -> mph
 
There seems to be an error in your calculations based on your assuming 241 Knots = 277 mph, which, however, is only true at sea level. At 25,000 ft altitude, 241 Knots = 403 mph (see my last reply above). Using the sea-level conversion for the Navy jet results in an underestimate of the circumference of the circle it travels, which you give as

((277/60)/60) * (360/1.4) = 19.8 miles

but which should instead be

((403/60)/60) * (360/1.4) = 28.8 miles

That larger circle may place the UFO further away. I'm not trying to be a nitpicking nag, just wanting to nail down exactly what the Navy jet's path was, because we can only know where or if the LOS angles (54˚L and 6˚R) intersect if we know the Navy jet's precise path. And it seems to me that it traveled about 3.4 miles around a circle with a circumference of about 29 miles.

There's another assumption in your above analysis that bugs me, which is: "If we take the target position as essentially fixed..." But what if it was not fixed? That assumption seems to set the UFO's position from the start, yet I want to find out what that position was, if possible. I suspect the best that can be done is to describe a range of possible flight-path scenarios as opposed to one definitive scenario. That the target was essentially fixed may be just one possible scenario.

I totally agree we could make some reasonable hypothesis. @Mick West could you show your mathematical model for the static UFO hypothesis updated with the correct turn radius? It would help visualise ranges I think.

The pilots say: "they are all going against the wind. The wind is 120 knots to the west". Could we use that to make another "moving UFO" scenario?
Could we use the brightness of the clouds to estimate where the sun is and therefore where "the west" is?
Given the brightest part of the clouds is at the bottom I would say the sun is probably behind the observer (or the top would be illuminated) so that would give us a general "south" as behind the observer and the left side of the screen would be west? So the apparent movement is somewhat confirmed by the pilot's statements. We should simulate the movement in that direction.

Also the 120kts will drift the F-18 sideways making its trajectory in real space more of a spiral rather than a circumference but I guess we can ignore that as the wind is uniform on both the object and the F-18.
 
So this guy on twitter from a relatively new account (April) is claiming the Gimbal video does not show glare as explaining the rotations.
But I do not understand what he is trying to say about the pod movement proving this.
Anyone know what he is talking about?
He seems to think that if the Gimbal was glare it would move smoothly as the heading changes gradually.

It's an understandable misunderstanding. I really don't discuss that aspect enough.

The sudden movements are from sudden gross movements of the Gimbal as it traverses 0° (and not AT 0°). They are minimized they are noisy. The actual fine tracking is done with internal gimbaled mirrors.
 
He seems to think that if the Gimbal was glare it would move smoothly as the heading changes gradually.

It's an understandable misunderstanding. I really don't discuss that aspect enough.

The sudden movements are from sudden gross movements of the Gimbal as it traverses 0° (and not AT 0°). They are minimized they are noisy. The actual fine tracking is done with internal gimbaled mirrors.
Just to be clear Mick that's an hypothesis right? We have no conclusive evidence that's how ATFLIR would actually behave in those conditions.

For all we know both the ATFLIR and the object might be rotating as we have no way of determining or understanding the actual movements of ATFLIR.

The Navy does have the ability to conclusively prove this obviously with a few tests on an actual ATFLIR of the time.
 
In any case: the pilots commentary is very precise and gives very clear and direct testimony of the data they had on their screens.

"There's a whole fleet of them look on the SA" (SA = situational awareness screen with data based on radar and data link from AWACS and other aircraft in the area through Data Link used to point ATFLIR)

"They are all going against the wind. The wind is at 120 knots to the west." (the F-18 has a screen calculating exact wind speed at that altitude)

We can conclude:
  • They are pointing the pod at one contact out of many (see this tutorial on how ATFLIR is used to point at contacts on the radar screen: )
  • the contacts are in controlled and sustained flight at around 25k feet
  • the contacts are able to achieve more than 120 knots at that altitude (not sure if the knots they refer to are indicated IAS or TAS) and move independently from the wind.
  • the contacts emit a lot of IR radiation and are detectable by radar (or would not be in the SA page)
  • the contacts are not US military jets or the SA page would identify them as friendlies
A couple of extra data points:
  • the symbology of the POD indicates it is in Point Track mode and the target is not "designated" (see:
    Source: https://youtu.be/HJRb_ofEtYQ?t=367

    for a nice schematic on the different symbols in a sim of this aircraft)
  • I'm not sure if F-18 has radar contact with the thing (no range indicated, target is not designated). Data on the SA page could come from some "donor" such as AWACS or the fleet radars.
  • See also this tutorial on how ATFLIR is guided by the plane radar:
    Source: https://www.youtube.com/watch?v=ChsMcv46Rj4
 
Last edited:
Just to be clear Mick that's an hypothesis right? We have no conclusive evidence that's how ATFLIR would actually behave in those conditions.

For all we know both the ATFLIR and the object might be rotating as we have no way of determining or understanding the actual movements of ATFLIR.

The Navy does have the ability to conclusively prove this obviously with a few tests on an actual ATFLIR of the time.

It's mentioned in the patents IIRC

https://www.metabunk.org/threads/gimbal-lock-and-derotation-in-flir-atflir-systems.10792/
 
In any case: the pilots commentary is very precise and gives very clear and direct testimony of the data they had on their screens.

"There's a whole fleet of them look on the SA" (SA = situational awareness screen with data based on radar and data link from AWACS and other aircraft in the area through Data Link used to point ATFLIR)

Worth noting that, per Ryan Graves, the 'fleet' refers to the 'little guys' - the types of contacts they'd been seeing regularly, and had been observed once visually as a 'cube inside a sphere' object.

Lt Graves says the Gimbal video is just part of a longer, higher resolultion video he was shown on the ship. It included several smaller craft, the kind he was more familiar with.

GRAVES: They were essentially flying themselves in a formation. And then behind those was the object you see here.

GRAVES: At this time is pretty much where you see this video, where these are now out of the picture, and it's just this guy that's stationary.

ELIZONDO: Is that why they refer to them as "there's a whole fleet of them"?

GRAVES: Yep. So there wasn't a whole fleet of these, there was just all these little guys. And there's the one object which we had never seen before, which was significantly larger.
Content from External Source
 
Might also be worth mentioning I looked into the DCS F/18 simulation and it seems the current simulation of the ATFLIR pod might not be sophisticated enough try and recreate some what we see from the 3 Navy videos GIMBAL especially.
 
Last edited:
  • I'm not sure if F-18 has radar contact with the thing (no range indicated, target is not designated). Data on the SA page could come from some "donor" such as AWACS or the fleet radars.
  • See also this tutorial on how ATFLIR is guided by the plane radar:
    Source: https://www.youtube.com/watch?v=ChsMcv46Rj4

In the video tutorial above it seems the "SLAVE" indicator should show "L+S" as selected if the ATFLIR was correlating its lock to a radar track. We don't see that in the gimbal video so I guess that points to the F-18 not having radar lock on the object at the time. Just a point lock with ATFLIR. It's not easy to manually slew the pod to track a moving and distant object (obviosuly). If it was in the SA page it was probably ATFLIR locked automatically based on that.

The F-18 might still have radar contact from its on board sensor however or get the data from donors through data link. Achieving a radar lock with the F-18s radar might be regarded in some cases as an hostile/dangerous behaviour and send off warnings in enemy/friendly military aircrafts.
 
Not being locked was mentioned in Micks interview with the anonymous avionics engineer, not sure if this the same vide being referenced.

https://www.metabunk.org/threads/av...flir-targeting-pods-tracking-and-glare.11392/

"K: Okay, so he's just doing an optical track on this as well. So he initiated that range in TV, just doing the optical track, it looks like here, so you are not going to get a range in TV, and he's not going to designate it a target obviously. Because I'm sure he knows, it's just a plane, airliner."
 

Source: https://twitter.com/uncertainvector/status/1394761778765221889?ref_src=twsrc%5Etfw%7Ctwcamp%5Etweetembed%7Ctwterm%5E1394761778765221889%7Ctwgr%5E%7Ctwcon%5Es1_c10&ref_url=https%3A%2F%2Fwww.redditmedia.com%2Fmediaembed%2Fnfyc3j%3Fresponsive%3Dtrueis_nightmode%3Dtrue


Ryan Graves was not a WSO/Pilot for GIMBAL right, the statements in this thread saying that GIMBAL is from unknown pilots etc are still accurate? Or has there been some new information since?

Also worth noting an update from new knowledge of ATFLIR SLAVE that the ATLFIR in GIMBAL is optically tracking, ie not slaved to L+S (primary radar track) or an other RADAR track.
 
Last edited:

Source: https://twitter.com/uncertainvector/status/1394761778765221889?ref_src=twsrc%5Etfw%7Ctwcamp%5Etweetembed%7Ctwterm%5E1394761778765221889%7Ctwgr%5E%7Ctwcon%5Es1_c10&ref_url=https%3A%2F%2Fwww.redditmedia.com%2Fmediaembed%2Fnfyc3j%3Fresponsive%3Dtrueis_nightmode%3Dtrue


Ryan Graves was not a WSO/Pilot for GIMBAL right, the statements in this thread saying that GIMBAL is from unknown pilots etc are still accurate? Or has there been some new information since?

Also worth noting an update from new knowledge of ATFLIR SLAVE that the ATLFIR in GIMBAL is optically tracking, ie not slaved to L+S (primary radar track) or an other RADAR track.

Graves is making claims on Twitter about this video was he the WSO for the GIMBAL video?
 


Huh...

I am sorry we don't have any video of all these other objects flying around, or the moment when the Gimbal object executed its physics-defying turn described in (3/6) or it joining the formation of objects in (5/6).

Am I correct that he is either involved in operating shipboard radar during the incident, or is speaking for a friend of his who was? Or, at least, that he's speaking about radar observations? Or am I misreading? He seems to be mashing up radar and visual observations, and confusing the issue by not making it clear which is which.

In any case, when he talks about the object rotating, when it clearly did not (those light artifacts inside the camera that rotate along with it, see post 361 of this thread among others, clench that argument) he loses a lot of credibility.
 
Am I correct that he is either involved in operating shipboard radar during the incident, or is speaking for a friend of his who was? Or, at least, that he's speaking about radar observations? Or am I misreading? He seems to be mashing up radar and visual observations, and confusing the issue by not making it clear which is which.
It would make sense that he has seen the radar tapes (which are always classified) from the aircraft involved in the encounter, as well as whatever camera footage they shot.
 
I am brand new to this forum so please remove this post if inappropiate. I talked about the perception of one of the radar operators that states he saw a fleet of object with equal spacing move from one side of the screen to another with a neighbour that was a captain on a commercial transport vessel. He gave me his old radar training books and some explanation that could clarify that sighting. Before i post a hypothisis i would like to talk a little bit with someone about it before i post. Is there a metabunk discord or voice app that i can just talk to people to check if i got the right conclusion? (i hate typing) :p

Edit: Im talking about Kevin Day's description of events i think, maybe the wrong topic
here?
 
Last edited:
Seeing again this gimbal video, why is the whole image frame jumping slightly at the exact same time as the rotations happen? This indicates to me that: either the camera makes a move at the exact same time the object rotates three times, or the rotation of the object is caused by the gimbal opto-mechanics itself. The latter is of course what I think is happening.

In case this has been discussed already, ignore my comment.
 
Seeing again this gimbal video, why is the whole image frame jumping slightly at the exact same time as the rotations happen? This indicates to me that: either the camera makes a move at the exact same time the object rotates three times, or the rotation of the object is caused by the gimbal opto-mechanics itself. The latter is of course what I think is happening.

In case this has been discussed already, ignore my comment.

Not only has this been discussed, this basically Micks entire point.


Source: https://www.youtube.com/watch?v=VuSKFwhXhoY


The object is a IR glare and rotates at the point where the camera has to physically rotate due to the gimbal mechanism, the horizon is corrected for by the de-rotation mechanism.

There's probably a reason the Navy called it GIMBAL internally.
 
The object is a IR glare and rotates at the point where the camera has to physically rotate due to the gimbal mechanism, the horizon is corrected for by the de-rotation mechanism.

There's probably a reason the Navy called it GIMBAL internally.

I remember this discussion, as i also actively hinted to the notion of the glare being produced by the optics, that can rotate around the optical axis.

What I mean is that the "bump" of the whole frame is an indicator for it being caused by the optics/glare. Just wanted to point that out..
 
The bump is the less than smooth transition as the Gimbal rotates over the top and has to adjust, unless you mean the other optics/glare the "light field" i.e. other glare type artifacts also rotating that proves it's not the object itself
 
What I mean is that the "bump" of the whole frame is an indicator for it being caused by the optics/glare. Just wanted to point that out..
I must have watched that gimbal video being picked apart and analyzed a hundred times, and never really picked up on the "bumps". They're incredibly obvious now you've mentioned them.

Even looking at a tiny part of the horizon is enough to see three distinct perturbations, and yes they happen right when the apparent rotations begin.

bump-out.gif
 
Hi all,

What a fantastic forum this is. I have a couple of points regarding the gimbal footage and the rotation/glare hypothesis, which continue to bother me so was wondering if someone could take a look.
In Mick Wests’ great gimbal video where the basic mechanics are explained really well, he says - “ the gimbal does a large rotational correction around the centre 0 deg mark “ This video is such a clever analysis but I don’t feel it quite tallies with the data on the atflir system screen and the mechanics of the gimbal.
We know from the gimbal footage on screen data that -
The camera starts looking 54 deg left and is constantly looking 2 deg down.
The F18 is in a roughly 20 deg left bank from the start.
The glare starts to suddenly rotate around 6 deg left point.
So we can see that it is a relatively steady slow and consistent path across the camera. I think this is critical for the glare hypothesis.
Ive made a very basic video in my workshop which shows that a subtle combination of the two shifting axis is required to track an object with no large corrections across that path. I can post if needed.
If we know that the camera is constantly working on the two axis to keep the craft centred then we can deduce that any glare must visibly be rotating consistently through the footage. It’s not possible that the atflir can track the object all the way across from 54 degrees left without any combination of tilt and rotation almost like a static camera held in hand, and then suddenly the rotation starts in the middle and the glare then rotates.
All the other footage on these forums and elsewhere have that constant- the rotational glare is continuous when filmed from a moving vehicle with a gimbal. Only if the craft was directly behind at 0 degrees level and 0 degrees left / right would it be possible to keep the glare from rotating. The only other time we see this is the handheld footage from Dave Falch and others which don’t have rotating glares because it’s a handheld camera without any shifting of lenses or mirrors. As soon as you add moving parts into the equation, you immediately would see the glare rotate as in Mick West’s own infrared lense rotation experiments.
Think of it like this -
If we tracked a distant bird flying steadily from left to right but also coming towards us at a 45 degree angle and then moving away from us as it reaches a centre point at an opposite 45 degree angle. The way we would track it with binoculars can be imagined as kind of similar to the mechanics of a gimbal. We would slowly move the binoculars left to right ( first axis ) whilst also rolling the zoom ( second axis ) in unison to keep the bird nicely centred. Then as it passed the centre point we would keep moving to the right but slowly zoom out again in the opposite direction. It’s the steady combination of the two movements which we need to keep track. Now if we imagine that when we use the zoom in conjunction with turning it causes a glare from our bird to rotate. Then we can see that, as long as we kept the bird in view, we would be creating a constantly rotating glare, slowly throughout the entire path. The rotation may be more exaggerated around the centre point but it would be visible throughout. The only way we would get just a sudden rotation in the middle would be if we lost track completely, then captured it again near the centre line, adjusted the zoom frantically to re acquire our target. Or if we didn’t use the zoom but ran backwards frantically away from the bird but kept it in view, then stopped as it reached the centre line and started using the zoom instead because we are exhausted ! Hope this makes sense ? It’s obviously not exactly the same as gimbal mechanics but it has two axis which must be used together to track a target which by virtue of turning in unison will create a glare rotation.

In essence, it seems to me impossible for the glare to rotate just at that point, when taking all the other data and movement into consideration.

Would like someone’s input on this . Many thanks.
 
I think the main front section of gimbal which rotates and tilts the lens does most of the leg work constantly tracking an object. Obviously when something is miles away and the camera is zoomed in (x2 in the gimbal footage for example) then the rotational motion of the front section cannot track finer movement within a frame. This is where the coelostat mirrors will do the work. Kind of like a different, larger gear on a bike that’s able to rotate to a much finer degree. If the object travels all the way across from past 45 deg left to in front of camera in a relatively smooth but short timeframe then I think all of the components would be constantly moving in unison. I don’t believe the coelostat mirrors could do all of that work. It’s too much of a span. The mirrors could only track the object if the front rotating/tilting section has it already in shot. If the object hung out on the left and just drifted from say 54 degrees to 50 degrees then I guess the mirrors could do that on their own. But if even the mirrors managed to track all the way and then a sudden rotation is required at centre it still doesn’t explain why the glare rotates only at that point. Regardless of whether just the mirrors are rotating , or the front lens or everything together ( which is what I believe is required ) the rotation of the glare would be constant, would it not ? Sorry if I’m missing something obvious here that has already been explained elsewhere but I can’t get my head around this point.
 
Last edited:
Back
Top