Agent K
Senior Member
The video is cut off at the most interesting point.The interesting thing about the Gimbal video is what it doesn't show.
ie TV mode footage. Why? Does it give it away?
The video is cut off at the most interesting point.The interesting thing about the Gimbal video is what it doesn't show.
ie TV mode footage. Why? Does it give it away?
ExactlyThe video is cut off at the most interesting point.
This is something that should be checked with raw data, and not in a frame extracted from a video... but...That's a really neat demonstration. This plus our Geogebra analyses are very consistent, we are making great progress here ! Now I'd like to understand how a Navy pilot can miss he is tracking another jet only 10-15Nm away.
I think it will be interesting to see what happens, but relying on the pod simulation to emulate reality to that level should be done carefully.
I'm kind of interested in what blind spots the pod has. Reading around makes me think it's main use is for ground target designation for laser guided bombs. The AA mode must require you to fly in a certain way to give the pod visibility of targets at some aspects. Interesting to see what happens when you ask it to track a target that moves into a blind spot.
Excellent work @MclachlanM
In the video the object clearly becomes significantly bigger at the end of the video compared to the beginning. In this simulations with an F-18 flying away are the relative distances reduced between the beginning and the end of the video accordingly in order to justify this increase in apparent size? Otherwise the flight profile does not match the observations.
I see the object start at 25x25 pixels at the beginning to 24x35 at the end. Quite an increase in apparent size especially in width (rotated).
From what I've been reading, full IR modeling is going to be in an upcoming big update. It'll be interesting to revisit these results then.The more I read on the DCS ATFLIR system the more I am convinced that it is actually attempting to model the actual IR radiation: I've heard that developers have spend years on the ATFLIR update ensuring that the IR simulation is done correctly on some forums, also IR is super important when modelling heat-seeking missiles, flare counter measures etc. so it makes sense if they have attempted to get as close to the real thing as possible. The fact that the range increases and target bars encompassing the entire object get smaller makes sense, however the effect of positioning behind the object, looking directly down the nozzle and increasing the bright IR size is astonishing to me.
Where do you get that understanding?The "gimbal lock outside rotation in limited movements" hypothesis is only suggested by the patent?
- how do we know the patent is related to ATFLIR?
- how do we know that ATFLIR actually implements the patent?
My understanding is that the inside mirrors are only used for fine tracking. Not for major tracking such as a large sweep across the bow. I would expect the ATFLIR head to rotate/move across the entire sweep and the small mirrors to do the fine track. Exactly like shown in DCS.
It's the most simple and reasonable system. We are tracking objects at extreme range and need to smooth out a lot of noise (e.g. turbulence etc.). It would be impractical/impossible to do fine tracking with the main gimbal system. You need something precise and fast for precision tracking and something else for the main large tracking. I would assume the internal mirrors constantly work to track the object within less than +/-10% of the window FOV while any tracking beyond that or even if you get close to the edge of this limit the main gimbal would recenter.Where do you get that understanding?
The patents describe something that looks exactly like the ATFLIR system from what we know. The Gimbal roll near 0° is something we see actually happening in the FLIR1 video, and there are examples in other systems where it's explicitly displayed on the screen.
It also makes a whole ton of sense that the main axis rotation would be minimized. The forward pod weighs probably over 50 pounds. You see the vibration of the scene at the start of the large rotations in GIMBAL.
Assuming we are seeing the object, maybe -- but I don't think we are. It might be possible to calculate the "size" of the glare, which is a pretty nonsensical statement but would give an upper limit on the size of the object (it must be small enough to be hidden by the glare.)We are ignoring one of Lehto's points though: the size of the object can be calculated/approximated by looking at the apparent size of the IR picture.
Assuming we are seeing the object, maybe -- but I don't think we are. It might be possible to calculate the "size" of the glare, which is a pretty nonsensical statement but would give an upper limit on the size of the object (it must be small enough to be hidden by the glare.)
the size of the object can be calculated/approximated by looking at the apparent size of the IR picture.
It can't be done with only the numbers in the video, you need a range which is what Chris is disagreeing with.How? Describe this mathematically.
We have concluded that the IR signature of a jet engine can obscure the aircraft when seen directly from the back but does not expand that much beyond that.If the engine is a bright IR light source, then it can be seen at any distance regardless of the objects size. The brighter the light, the bigger the glare.
I think a pilot's intuition given their experience will unconsciously put together the 5 points above so the pilots have probably a better idea of what we are looking at at first glance compared to us.I don't think anything less than a good bit of computational geometry is actually going to come up with the range/estimate of what we're seeing in the Gimbal video.
And that all comes back to the title of this thread. Does Chris Lehto absolutely know what this object is doing based on this footage. No freaking way. I can guarantee any calculations based on three lines and three headings isn't going to get you there. This thing could be stationary or traveling at 2000 m/s... it's all in some very touchy math and even if we can get some numbers there's going to be a range between balloon and supersonic.
We have concluded that the IR signature of a jet engine can obscure the aircraft when seen directly from the back but does not expand that much beyond that.
So assuming it would loosely match the wingspan of the aircraft in size is reasonable.
That is the smallest glare size that's compatible with the hypothesis, but it could be larger.So assuming it would loosely match the wingspan of the aircraft in size is reasonable.
How much larger?That is the smallest glare size that's compatible with the hypothesis, but it could be larger.
You can see it doesn't look anything like those IR signatures.
Arbitrarily so -- a bright enough IR source on the other side of the universe, with completely negligible angular extent, could still produce a glare of measurable size on a camera lens here. That is also one problem with the attempt to estimate distance to the object using the apparent size difference of the black shape: you don't know how much of the size difference is because the source appears brighter because it's closer (and thus resulting in a larger glare) and how much is because the angular size itself of the source is larger. You'd need it to be 100% the latter to come up with a useful estimate, but it's more likely to be mostly the former. Without access to the exact ATFLIR pod that was used, that exact day, in order to characterize its point-source response, we have basically no insight on the explanation for the change in apparent size of the black shape.How much larger?
The Chilean navy video was shot from a fairly low altitude, and Dave Falch's videos were shot from the ground. It's not clear to me that the flickering has anything to do with the operation of the engines; it could just as well be due to atmospheric turbulence. Compare Falch's videos with the images here, for instance. The air higher up in the atmosphere is typically calmer so it's not surprising we don't see any of that in the gimbal video.I'm not convinced what we are seeing is "glare from an engine". Just like it's not glare from an engine in the chilean UFO video. I've rewatched it a couple of times and you can clearly see it flickering
The size of the original IR source is not what's relevant, its brightness is. As long as it's bright enough it can appear as large as it pleases.In any case. As the video above shows: the glare of the engines rarely covers the entire plane. Si it is improbable that it will be much larger than the original IR source.
By exposure I think I mean "level/gain" in terms of ATFLIR settings.It looks exactly like those IR signatures?
You need to bear in mind we are seeing completely different equipment and that the settings/image may differ, but it is remarkably close for it to not be glare.
Edit: you can get a feel of how different settings in the ATFLIR may effect the image here - https://www.metabunk.org/threads/gimbal-id-of-a-possible-jet.11828/post-252110 , not sure which one you mean when you say exposure?
Isn't brightness directly correlated to the angular extent?Arbitrarily so -- a bright enough IR source on the other side of the universe, with completely negligible angular extent, could still produce a glare of measurable size on a camera lens here. That is also one problem with the attempt to estimate distance to the object using the apparent size difference of the black shape: you don't know how much of the size difference is because the source appears brighter because it's closer (and thus resulting in a larger glare) and how much is because the angular size itself of the source is larger. You'd need it to be 100% the latter to come up with a useful estimate, but it's more likely to be mostly the former. Without access to the exact ATFLIR pod that was used, that exact day, in order to characterize its point-source response, we have basically no insight on the explanation for the change in apparent size of the black shape.
Yes this could be the reason. We do see the same flickering when watching "normal" images so I guess this is a reasonable explanation.For example, say you're looking at a bright IR source such that (ignoring glare) it would appear exactly 1 pixel wide in your detector. You halve the distance to it, and now it would be 2 pixels wide in your detector. The diameter of the glare can only change by 1 pixel due to this effect. However, now the IR source appears 4 times brighter, which could increase the diameter of the glare well in excess of that 1 pixel. Without characterizing the camera in detail you can't extract that 1 pixel from the glare noise.
The Chilean navy video was shot from a fairly low altitude, and Dave Falch's videos were shot from the ground. It's not clear to me that the flickering has anything to do with the operation of the engines; it could just as well be due to atmospheric turbulence. Compare Falch's videos with the images here, for instance. The air higher up in the atmosphere is typically calmer so it's not surprising we don't see any of that in the gimbal video.
Those are high bypass engines (so a lof of "ambient" air being pushed. And I'm pretty sure the fact we don't see the full exhaust depends on the level/gain of the camera. How else do you explain the image I posted in my comment when the chilean navy gain/exposure is changed and the full exhaust appears?Note also that if what we're seeing in the Chilean navy case were in fact exhaust gases, we should see the exact same effect on the ground, during takeoff. If anything it'd be stronger (higher throttle setting). But you don't see that. On that basis it seems pretty conclusive that it's glare.
Agree. But the brightness of a typical jet engine is (somewhat) known. In most cases we can see it doesn't extend beyond the wingspan (often it is smaller actually).The size of the original IR source is not what's relevant, its brightness is. As long as it's bright enough it can appear as large as it pleases.
That's motion blur from the camera FOV change. You see it in other placesI'm not convinced what we are seeing is "glare from an engine". Just like it's not glare from an engine in the chilean UFO video. I've rewatched it a couple of times and you can clearly see it flickering and in certain moments where the exposure is changed you can see the full extent of the jet exhaust IR signature visible.
If the footage is showing the auto level and gain then ALG will be boxed on the DDI, what we see is the WSO's own settings for level, gain and contrast. So not necessarily better than what Falch is using (also the plane is much further away).But I would assume ATFLIR would be better than what Falch is using. Expecially in terms of autofocus/autogain etc.
It doesn't have to extend beyond the wingspan as the level, gain and contrast settings can easily pick out the IR glare and make the rest of the plane blend in with the background.Agree. But the brightness of a typical jet engine is (somewhat) known. In most cases we can see it doesn't extend beyond the wingspan (often it is smaller actually).
That's motion blur from the camera FOV change. You see it in other places
That image confirms my hypothesis: if you measure the width of the "glare" it approximates the wingspan/lenght of the original aircraft. It doesn't extend much beyond the actual fuselage of the aircraft. Of course it's an approximation. But a pretty good one. We know the error might be around 100% max? But from the videos I've seen so far not 5x or 10x. Jet engines are just not bright enough.If the footage is showing the auto level and gain then ALG will be boxed on the DDI, what we see is the WSO's own settings for level, gain and contrast. So not necessarily better than what Falch is using (also the plane is much further away).
It doesn't have to extend beyond the wingspan as the level, gain and contrast settings can easily pick out the IR glare and make the rest of the plane blend in with the background.
From that video:
Whatever settings Falch is using make the plane invisible and you can only see the glare. This is also true of the ATFLIR system.
That's the contrail (ice crystals forming as the exhaust freezes. It's interesting how it "appears out of nowhere" as warmer than the environment. I suppose the chemicals in the air would emit less IR than the organized H2O crystals in the contrail? Weird.I agree. Although IR can show exhaust but it looks a lot different:
My point is we can use the size to approximate range.I'm confused as to what the argument is here?
This is clearly a different effect than what we see in GIMBAL?
Doesn't look like motion blur though. It isn't a straight projection of the original image (like your cloud example). It has a shape that changes and expands and follows the expected exhaust don't you think?That's motion blur from the camera FOV change. You see it in other places.
And here it's seen in the clouds
No, it's at the exact same angle as all the other (numerous) camera-change streaks. And there's a variety that don't look like simple motion blur. Probably because it's not perfectly linear motion.Doesn't look like motion blur though. It isn't a straight projection of the original image (like your cloud example). It has a shape that changes and expands and follows the expected exhaust don't you think?
Yeah. It is a follow up of a follow up of a follow up.. It gets a bit boring, but that is perhaps also because we know the story so well by now.The ad hominem attacks through Youtube videos are getting embarrassing, on both sides...
Herd to fit a couple of ads into that, though...42 seconds, rather than 1279 seconds.
The FOV he uses is correct, but to get the object size he needs the range.2. I am pretty sure their FOV and object size calculation are off.