Curious George
Active Member
Quick dump of the focus
Skinwalker Ranch - Season 4 Episode 10 - 3600 MPH UFO Claim
- Thread starter Curious George
- Start date
Curious George
Active Member
I can calculate the circle of confusion of the fly, which can be translated to a radius of Gaussian blur approximately.
Curious George
Active Member
Who blurred the video?
The viewer software has a built-in blurring function:
@MickWest - Try to apply a radius 3 or radius 5 Guassian filter to the YouTube sourced, clear, episode preview version of the fly and see if the result equals the production version.
Edit: I think it unlikely this was filtered directly in the camera software, but probably by production. Regardless, a Gaussian blur is pretty standard in video editing software. Also, if a larger blur is indicated (>=7x7), it's evidence that it was production that blurred it. In fact, if production has an original clearer version and made the YouTube post, they had to be the ones to blur it. Otherwise, they were given a blurred copy then got an unblurred copy for the preview.
The viewer software has a built-in blurring function:
@MickWest - Try to apply a radius 3 or radius 5 Guassian filter to the YouTube sourced, clear, episode preview version of the fly and see if the result equals the production version.
Edit: I think it unlikely this was filtered directly in the camera software, but probably by production. Regardless, a Gaussian blur is pretty standard in video editing software. Also, if a larger blur is indicated (>=7x7), it's evidence that it was production that blurred it. In fact, if production has an original clearer version and made the YouTube post, they had to be the ones to blur it. Otherwise, they were given a blurred copy then got an unblurred copy for the preview.
Last edited:
Curious George
Active Member
What Mick found was that the blur was a match for an object at 7-8 feet, which is close to one of my solutions at 7.8 feet, moving at (up to) 2.5m/s (could be against the wind) and 25mm in size.
Edit: the size will shrink based on newly noted evidence that the video was stretched.
And MIck if you are reading this July 4, first happy Independence day, and note that I spent the night editing the 2nd page of posts.
Edit: the size will shrink based on newly noted evidence that the video was stretched.
And MIck if you are reading this July 4, first happy Independence day, and note that I spent the night editing the 2nd page of posts.
Last edited:
Curious George
Active Member
22mm FL measured vs 28mm FL shown on lens
The 28mm vs 22mm FL confusion could come from the crop factor. What is the crop factor for a 20.48mm sensor?
The diagonal of the Photronic sensor is 28.96mm. The DX format is 24x16mm, with a diagonal of 28.84mm, almost the same.
I do note that 20.48/16=1.28 and 22*1.28=28.16, in other words if the equivalence between 22mm FL and 28mm FL on the lens can be explained by the difference in sensor heights, this would be a match.
Edit: This is a good example of trying too hard to fit a narrative. There's no reason to think the setting seen in the video remained during filming. Still, confirming this point should show that and the data will honestly say 22 is not 28.
To do: review definition of crop factor
The 28mm vs 22mm FL confusion could come from the crop factor. What is the crop factor for a 20.48mm sensor?
The diagonal of the Photronic sensor is 28.96mm. The DX format is 24x16mm, with a diagonal of 28.84mm, almost the same.
I do note that 20.48/16=1.28 and 22*1.28=28.16, in other words if the equivalence between 22mm FL and 28mm FL on the lens can be explained by the difference in sensor heights, this would be a match.
Edit: This is a good example of trying too hard to fit a narrative. There's no reason to think the setting seen in the video remained during filming. Still, confirming this point should show that and the data will honestly say 22 is not 28.
To do: review definition of crop factor
Last edited:
Curious George
Active Member
Meaning of "72.6%" zoom part 2
Confirmed, the operator is shown clicking "Fit" in the viewer software, and it's on, which is on the quick toolbar, and the size of the image is about 743x743 which is 72.6% of the full sized image.
Notice the mouse is captured and I can't imagine someone pressing a button to screenshot at that moment either. HDMI live feed?
Confirmed, the operator is shown clicking "Fit" in the viewer software, and it's on, which is on the quick toolbar, and the size of the image is about 743x743 which is 72.6% of the full sized image.
Notice the mouse is captured and I can't imagine someone pressing a button to screenshot at that moment either. HDMI live feed?
Last edited:
Curious George
Active Member
Frame rate of the series, frame rate of the slowed down video
I can confirm that this show is produced at the NTSC Film rate of 24/1.001=23.976 frames per second. That means that if the camera filmed at 2000fps and every frame was shown at NTSC Film rate, the resulting slow down is 2000/24*1.001=83.42 times. Regardless, each frame is 500uS.
The software can preview the video with a "skipframe" setting of 1, meaning each original frame is played, with a default playback speed of 30fps on the computer monitor, however it's clear that the file was given to production to edit for presentation. In that case, I assume the slowdown as above.
Edit: it was interesting that the software screenshot included the mouse at the moment of clicking the Fit tool. I can't imagine someone hitting a screenshot at that moment. Perhaps it was captured from live HDMI video, split to production. Can we see any wires?
The video can be saved as either individual frames, or an AVI/MOV/MP4 video file:
Although the video may have been saved as 30fps, you can tell that it wasn't by the smoothness of the playback, which shows that no frames were skipped or duplicated, given that the show is at 23.976fps.
To do: find a way to confirm the slow down amount. Could try measuring the rocket speed. Confirm no skipped or duplicated frames for at least 2x30-1 frames or whatever. If that's the case, the slow down can only be every frame or every 2 frames etc.
Edit: confirmed, all the slo-mo video footage is smooth.
Frame Action
3517 rocket, mid-frame, slomo
3646 end rocket slomo, near top
I can confirm that this show is produced at the NTSC Film rate of 24/1.001=23.976 frames per second. That means that if the camera filmed at 2000fps and every frame was shown at NTSC Film rate, the resulting slow down is 2000/24*1.001=83.42 times. Regardless, each frame is 500uS.
The software can preview the video with a "skipframe" setting of 1, meaning each original frame is played, with a default playback speed of 30fps on the computer monitor, however it's clear that the file was given to production to edit for presentation. In that case, I assume the slowdown as above.
Edit: it was interesting that the software screenshot included the mouse at the moment of clicking the Fit tool. I can't imagine someone hitting a screenshot at that moment. Perhaps it was captured from live HDMI video, split to production. Can we see any wires?
The video can be saved as either individual frames, or an AVI/MOV/MP4 video file:
Although the video may have been saved as 30fps, you can tell that it wasn't by the smoothness of the playback, which shows that no frames were skipped or duplicated, given that the show is at 23.976fps.
To do: find a way to confirm the slow down amount. Could try measuring the rocket speed. Confirm no skipped or duplicated frames for at least 2x30-1 frames or whatever. If that's the case, the slow down can only be every frame or every 2 frames etc.
Edit: confirmed, all the slo-mo video footage is smooth.
Frame Action
3517 rocket, mid-frame, slomo
3646 end rocket slomo, near top
Last edited:
Ravi
Senior Member.
OpticStudio it's called now..Excellent work Mick! This is truly a Sherlock moment
The sensor size remains 20.48mm.
The AoV is incongruous to the Google Earth estimate, this should be resolved -but in any case, the image is what matters, and these formulas I'm using are for a thin lens model, a simplification of a real lens. The only way to be better is if this model of lens shows up in Zemax (a popular optical simulation program).
The aperture estimate will be useful for calculating hyperfocal distance.
The cropping deduction of course comes from the software screenshot.
I am late to the party..
But finding the lens design of the Nikon objective used here, is I think not easy.
Also, I have not found the sensor-distance-to-mount values anywhere in the documentation of the camera. This is of course essential to know the FoV as the larger the distance, the more "macro" the shots will get. Perhaps there is even a distance ring in between the objective and the camera?
Here's the "rawest" comparison with the teaser and broadcast version. Teaser is the darker more grainy version.
[compare]
[/compare]
These are extracted directly from the 1080p YouTube files, NOT screen-captured. I saved the same frame losslessly in the video size of 1920x1080
Of course it's not the original raw file, which would be 1024x1024, and this particular clip is a bit less than the full width of the frame, so this is approximately double the raw file's resolution. But it's not an exact upscaling, so we don't see 2x2 pixels when we zoom it. But it's consistent with the basic doubling.
Ideally we'd see the raw 1024x1024 view, which presumably still exists somewhere.
[compare]
[/compare]
These are extracted directly from the 1080p YouTube files, NOT screen-captured. I saved the same frame losslessly in the video size of 1920x1080
Of course it's not the original raw file, which would be 1024x1024, and this particular clip is a bit less than the full width of the frame, so this is approximately double the raw file's resolution. But it's not an exact upscaling, so we don't see 2x2 pixels when we zoom it. But it's consistent with the basic doubling.
Ideally we'd see the raw 1024x1024 view, which presumably still exists somewhere.
Scaramanga
Senior Member
Their claim is utterly unscientific. This is classic UFOlogist BS....which annoys me intensely....as they cannot possibly say it went half a mile in 1.12 seconds without knowing the distance of the object. And they have absolutely nothing, no frame of reference, that provides that distance.
We simply do not know if the object was 10 miles away or 10 feet away. The fact that the object is blurred suggests it was close, as the camera seems to be set to infinite focus which is why distant objects are relatively sharp.
Sheesh....I cannot believe these guys have the utter gall to call themselves scientists.
Mendel
Senior Member.
They're showmen. They play scientists on TV.
Ann K
Senior Member.
Thanks. We've said all that before, but it needs to be repeated at regular intervals. We will, I'm sure, hear that kind of speed and distance claims again ...and again ...and again...
Curious George
Active Member
Hello Ravi!
Good point, however I don't believe that is necessary. The Focal Length and Angle field of View and sensor size relate to each other and nothing else. Mick has used image analysis to find that the AFoV was 50°, therefore the FL is 22mm, given that the sensor size is 20.48mm square. Note: 22mm is not the same as Equivalent (35mm, full frame) Focal Length which is marked on photographic lenses!
I could however, go into more detail.
Technically, this is the effective focal length, or optical focal length, for a thin lens model which is all that matters here. When you're talking about mechanical distances, you're possibly referring to back flange distance, this is from where the lens screws in to the sensor, and is defined by the lens standard. In this case, a Nikkon lens uses an F-Mount, and the flange distance is 46.5 mm. This is only a reference point to design the lens to. Then there is the back focal length; this is from the lens to the sensor. There's a few fine points here. The "lens" point would actually be the optical centre or nodal point of the lens, which is usually about mid-way through the lens, to the principal plane, which is where it would be focused when the lens is focused at infinity (which moves the lens closest to the principal plane). The idea is to use the construction of the lens, knowing the flange distance, to match the principal plane (where the image is focused) to where the sensor is. The back focal length is not necessarily from the last "bulge" at the inner end of the lens to the sensor, nor does the last part of a lens element necessarily correspond to the flange (it would seem prudent to have it a bit recessed so you don't scratch the inner end if you set it down).
I didn't mention more advanced aspects like; wavelength of light (each colour of light focuses differently, the standard "green" used which is the Hg e-line), lens distortion, and probably some other things.
Equivalent focal length is the focal length on a full frame camera. If the sensor is smaller than 35mm, you divide by the crop factor to get the real focal length.
If you used a simple convex lens, like a magnifying glass, the nodal point is exactly in the middle, and the back focal length and front focal length are the same.
There's multiple AoV; the horizontal, the vertical, and the diagonal. This is because a circle doesn't fit into a rectangle. For the sensor in the highspeed camera, the sensor is exactly square, so H-AoV and V-AoV are the same.
References:
Relating Angle of View to Focal Length:
http://www.artdecocameras.com/resources/angle-of-view/
Back Flange Distance
https://www.edmundoptics.ca/knowledge-center/application-notes/imaging/lens-mounts/
Last edited:
Curious George
Active Member
Watch the preview from 2:43. I believe the reasoning used by Travis, is that they were facing the camera and looking at the sky. Since they didn't see the UFO, he assumes that it passed behind him. He then estimates the visible length of the East Field as 1/2 mile.
Of course, this is only a rough estimation, and he doesn't even calculate that correctly, by using 1 mile in a simplified calculation.
It could be that a strong psychological bias is at play here. Further, other members of the team may be deferring to him as an authority.
Curious George
Active Member
This is only a technicality, but you can do slightly better. Notice the histogram:
The dark lines repeat every 7 pixels, which means the contrast was increased by 7/6=16%. This is a very telling number. There is a concept called "video levels" where the luminance, known as the "Y" in "YUV", is restricted to the range 16-235, or 220 levels. There are historical reasons for this I won't get into, but it gets converted to full range RGB which creates these gaps. Note that 256/220=1.16, or 16%.
I have a theory that you could invert the zoomed in version of the video, which could be using a higher precision of the raw video internally, to get back the full range values. Same with changes in contrast.
Another point, is that reversing the zoom back to the original size, would reduce any artefacts from the compression.
There is a more accurate way to screenshot such footage (in some sense), by mapping Y directly in RGB, which would make a dimmer looking image with no gaps, but it's a portable way to transport the actual luminance. Note that JPG uses YUV directly, but PNG/BMP/TIFF uses RGB.
Since there's no colour, the UV components are stored as 128 in a video file.
Reference:
https://www.thepostprocess.com/2019/09/24/how-to-deal-with-levels-full-vs-video/
Last edited:
Ann K
Senior Member.
I went back to look at that, and someone says "It went from halfway on the mesa to just over those trees in a second". In other words, they looked at it as if it were just a two-dimensional picture plane, and since it appeared (in that 2-D space) over the mesa and over the trees, they decided to treat it as if it were really traveling all the way from the mesa to the trees, when what it did was merely travel the distance to subtend that angle of view, at whatever distance it was. As @Scaramanga says, they can't calculate it without knowing the distance.
My personal thought is that it isn't a "psychological" bias. It's a bias in favor of marketability of their product, and their product is woo.
Curious George
Active Member
Yep, that makes sense.
I know a lot of people assume it's just fake, but that's not the impression I get from many interviews of people working on the ranch, plus guest contractors. Or, it could be only a few people are in on the "secret".
More than that, Travis talks on unrelated topics outside SWR where he makes the same errors and jumping to conclusions, such as his comments on the US Nimitz incident. An example, he called some horizontal lines in the footage after an optical effect, but I believe it was only interlacing, a concept from old video systems, as the footage was stored on basically VHS tape in the plane.
A lot of people don't realize, the Nimitz footage has extra blur in the horizontal direction only, and that a lot of strange artefacts can come from resizing interlaced footage and so on.
Last edited:
NorCal Dave
Senior Member.
Because that assumption is in line with the nature of the show.
If there was a fly buzzing between them and the camera, they may not ever notice it. They're out on a semi-remote plateau in rural Utah, there are probably all kinds of critters buzzing around. If one of them creates the illusion of a UFO, that's a bonus. It makes something that looks strange, and they can legitimately say they didn't fake anything, this is all natural.
IF they noticed a fly buzzing between them and the camera, then they would have to go back and look at the footage and make a decision. Do we go with it and say to our audience "this is just a fly"? Or do we admit it's just a fly and not show the video? OR do we just go with a fly buzzing and give our audience what they want, a UFO?
I imagine it's the former, but Taylor is plenty smart enough to know there is a relationship between size, distance and speed. In this case entertainment is more important.
You're trying to unravel it is commendable and I've been enjoying the discussion, though some of it goes over my head. I think it's important to debunk as many of these SWR videos as possible, if for no other reason to show it has been debunked. Taylor and Fugel don't care, but maybe someone will ask about it and there will be an explanation for the claims here.
Mendel
Senior Member.
Remember that fly=UFO has worked for them before, see this 2020 analysis: https://www.metabunk.org/threads/black-ufo-at-skinwalker-ranch-a-fly.11271/
Btw, this time, didn't they have IR as well?
Curious George
Active Member
I'm not really interested in discussing anyone's motives. I'm not here to accuse anyone, or attack anyone. Take the show as a form of entertainment if you like. I just find the video analysis interesting. Thanks.
Scaramanga
Senior Member
Well, personally I can't believe Travis Taylor is on the AARO team as his 'science' is about as pseudo as it gets. Given that the 'UFO' is almost certainly a flying insect of some sort, its not at all surprising that the team did not see it. To use the failure to visually see an insect as rationale that it must have been a mile or so away is just baloney of the first degree and the sort of 'logic' that not even first year science students would make....let alone someone given a position of responsibility by Congress.
But none of it should really surprise anyone, given the ability of the Skinwalker team to conjure up dino-beavers, 10 foot tall talking wolf-men, invisible 'anomalies' 5000 feet up......oh, and don't dig up the ground at all or the boogie man will appear. It is all par for the course.
On an aftershow "Ranch Insider" video Travis Taylor explained why he thought it was not a fly.
Breaking that down:External Quote:
False. Firstly it IS a bit blurry, and the experiments with a similar camera setup show you have a similar level of blur at 4-8 feet, not the 16 feet he suggests.External Quote:
Ludicrously false. A meter per second is 2.23 mph, lower than average human walking speed. Insects very obviously can fly much faster than that - just based on everyday experience. A house fly has an average speed of 5 mph, with a top speed of 15 mph. A bee flies at 15-20 mph. A horsefly has a top speed of over 35 mph (16 meters per second)External Quote:
False. We are in the realm of garbage-in/garbage-out at this point, as he was so wrong with the top speed of a fly. However that 10-20cm at 1 meter gives a super-wide field of view of 136° to 157°. Fitting the view in Google Earth, and doing the math on the camera and the lens gives a FOV of just 50°External Quote:
True in itself, but false in that it wasn't that close to the camera. Taking into account the blur level and the camera FOV that works out for something about 6 feet (2m) away, which also works perfectly for size and speed (about 5 mph) of something like a house flyExternal Quote:
Last edited:
NorCal Dave
Senior Member.
Not to get off topic, but Taylor was part of the UAP Task Force under Jay Stratton back in ~'18-'21(?) or so. I was not aware of him being part of AARO.
Curious George
Active Member
It would have been 4 feet away
|
|
|
|
|
| ||||||
|
|
|
|
|
|
|
|
|
|
| |||||
|
|
|
|
|
Most of the math here is very simple. The object measures 17 pixels. The video is HD with 1920 pixels across. That means 1920/17=113 lengths of the object across. This requires no assumptions whatsoever (but ignores lens distortion, which could change the distance a few %). So, if the object is 10 mm, the distance it travelled is 10*113=1.3 m. If you go 1.3m in 1.12s 1.3m/1.12s=1m/s.
The AoV was measured based on the zoomed in video, but it covers 86% less width of the original video (by Mick's estimation), meaning that what they saw cross the screen went further than what we saw - in fact, the first shot of it has 19 frames where we don't see anything before the object appears at the left side. That means we should increase the speed by 17%.
The formula for how far away it is only needs the angle of view, and it's still simple geometry. WD=Field/2/TAN(AoV/180*PI()/2). WD means Working Distance or how far away the object is. AoV means Angle of View (sometimes called Field of View). The 180*Pi is to convert degrees to radians. WD=1129/2/tan(50/180*pi/2)=1211.
This would require a Focal Length of 2-4mm. This is clearly not possible with the lens we saw in the video, nor any lens. At that point you need to shoot off a mirror ball.
Edit: shortest I could find was 4mm, this is called a fisheye lens. I didn't know much about them. They have severe distortion, you would know right away if it were being uised.
Last edited:
jarlrmai
Senior Member.
Yeah not a part of AARO that I know of, but as part of UAPTF it's possible Travis Taylors flawed analysis and subsequent spreading of that is in part responsible for the current UFO wave from Elizondo to Grusch and now the US Senate. All feeding back in on each other.
And here he is seemingly unable to do basic analysis.
Curious George
Active Member
To put that in perspective, Travis is reported to be 1.85m tall. So he's saying it would take almost 2 seconds to fly from his feet to his head. That's not very fast. Hindsight is always 20/20 though.
Curious George
Active Member
I have a lot more work to do in my analysis. Don't take anything I say as final as it's a work in progress. I will do a summary writeup when I'm done. It's not so much that I don't think it's a fly, or rebutting the counter-claims or justifications, it's a sample problem for me to develop a toolbox and experience in this type of analysis. I'm more interested in using this for a portfolio or something.
Actually pretty accurate, in that this typical fly is 8mm long, exactly the same as the diameter of the head of the screws that I used.
Curious George
Active Member
On 7/5/2023, on the Insider's live chat, Travis asked for events that could have a prosaic explanation in any episode aired to date.
Yesterday? Where was this?
Curious George
Active Member
Sorry, it was Ranch Revelations on Thursday 7/6, a regular 1.5 hour chat on the Insiders Website. Last week was just Brandon, this week Travis was on.
The recording hasn't been posted yet. Check in a few days.
They've been able to process the fact that it's a fly now. I think this is ultimately good. We might see a more scientific investigation now.
The recording hasn't been posted yet. Check in a few days.
They've been able to process the fact that it's a fly now. I think this is ultimately good. We might see a more scientific investigation now.
Last edited:
NorCal Dave
Senior Member.
I wouldn't bet a dozen donuts on that.
They have to walk a tight line between doing actual research and scientific investigation and entertainment, with entertainment being the main point. They can try to do "entertaining scientific investigations", something the Mythbusters were better at, but their audience ultimately wants weird stuff, not flies in front of the camera.
Taylor especially is aware that there are non-fans of the show that have shown how silly some of it is. @Steven Greenstreet did a rather brutal takedown of the whole show after Taylor recommended Fugle not let him anywhere near the ranch. Others here, you now included, have shown that what they claim is often misleading at best and total nonsense at worst.
Called out enough for completely ignoring the possibility of a fly, Taylor explained why it was not a fly. An explanation that may have worked for fans of the show, but which Mick showed to be mostly wrong (post #62). I don't know if Taylor reads this forum or not, though I imagine he's aware of it. The fact that they are now saying it was a fly is a result of pushback from people like you, showing how wrong they were.
When it became too obvious it wasn't a UFO, they pivoted. But they're 4 seasons in now with a spinoff, and at some point, they're going to face the Ancient Aliens decision. Do they continue with a veneer of scientific plausibility and occasionally admit a UFO was a fly? Or do the just ignore the debunkers and skeptics and go full Monty with weird stuff, no matter how silly. AA has been recycling the same old fake and debunked material for over 10 years and their core audience still watches.
Mendel
Senior Member.
Flapping Drones
Source: https://m.youtube.com/watch?v=aj3wxJzrofI
Article: Mostafa Hassanalian, PhD
Associate Professor, Mechanical Engineering
Research Areas:
Biomimetics, and Bio-Inspired Aerial and Aquatic Robots
Design, Optimization, and Performance Enhancement of Aerial Vehicles,
Drones (UAV/MAV/NAV/PAV): Fixed Wings, Flapping Wings, Rotary Wings, Tilt-Rotor and Tilt-Wing Drones, Morphing Drones, Space and Marine Drones, Separation and Swarming Flight of Micro Drones.
New Concepts for Planetary and Space Exploration
Aerodynamics, Hydrodynamics, Aeroelasticity, Fluid-Structure Interaction
New Mexico isn't all that far from Arizona...
Ann K
Senior Member.
Just watching the vibrations set up by flapping wings makes me think that structural integrity would be a serious problem. Are those things just experimental novelties? Any helicopter-like rotary blade apparatus would give a much more stable flight, wouldn't it?
NorCal Dave
Senior Member.
No, but New Mexico Tech it is about 500 miles from the area of Skinwalker Ranch my friend. Not sure they have that much battery power.
They are interesting, but as @Ann K said above, it seems they would just vibrate apart from all the back and forth. I wonder if birds have the same reaction humans do to things that are almost human but not quite. We find them creepy and disturbing. I've watched ducks and geese swim around plastic decoys of themselves, so I suppose not.
It is a hugely simplistic representation of bird flight. It does not reproduce any of the actual fine wing movement, Examples of wing movement in slow motion (full video included as various examples throughout), nor tail shape(itself a significant aspect of bird flight). Birds wings are not the fixed flapping aerofoil presented in the model but make adjustments in wing shape throughout the stroke/beat.
Anyone who has watched kestrel (Falco tinnunculus) hovering in a wind making slight adjustments in wing shape without flapping Kestrel hovering (from 0.01)
, or a peregrine falcon (Falco perigrinus) in a stoop at 200mph with wings almost closed, but again making slight adjustment peregrine stoop (from 3.05)
, will have noted just how complex bird flight actually is.
Similarly, there are different strategies in terms of wing beats themselves. Manx shearwaters (Puffinus puffinus) have a rapid stiff winged beat followed by long glides close to the sea surface Manx shearwater (from 000) (I have watched them off Walney Island and this video does not show their true ability, where they will glide in troughs between waves inches above the surface).
The flappy model may appear bird-like but the flight is a poor representation of actual bird-flight as it lacks the nuanced changes of wing and tail shape. ( by accident the model does reproduce the undulating flight of eg Woodpeckers/Finches but they do not use constant wing beats to produce it as the model does).
'Experimental novelty' gets my vote.
JMartJr
Senior Member.
I've had birds come in to observe more or less bird shaped kites, and sometimes to attack them. And, on one odd occasion, to sit on one! Other than that last one, they seem to react as they would to an actual bird -- as far as I can decipher bird intentions.
Ann K
Senior Member.
And there's the tried-and-true method of putting a cutout of an owl on a glass door to stop small birds from flying into them to attack their own reflections. It works for me! Many tall structures like church bell towers put plastic owl models there for the same reason: birds see a predator and avoid it.
Similar threads
Trending content
-
-
Thread '"Active Social Engineering Defense" and "Large Scale Social Deception"'- Mick West
Replies: 22 -
-
-
