There's a few highly suspect details in the way the author got the 34-35 degree estimate. Maybe he was just really lucky and the errors cancelled out, but don't count on using his method to produce similar results.
Polaris is about .8 degrees away from the celestial North pole, so using it alone can give a reading off by .8 degrees, and the orientation in the photo is actually nearing the worst case.
The horizon doesn't appear in the photo. I have done celestial navigation strictly as a hobby for fun, and using a sextant without a visible horizon results in a few degrees of error at best. And that's with the benefit of the sensation of gravity helping you estimate. It is theoretically possible to estimate the horizon using the man made parallel horizontal lines in the image, but the author doesn't give any indication he did that.
Additionally, the author used the angle measurement tool in Stellarium to measure the distance to another ambiguous horizon. This is incredibly odd, as just clicking on Polaris in Stellarium will tell you its elevation. But even more odd, is he would have had to explicitly set his latitude in the app in order to get that screenshot.
I'm sure you're correct, as the whole thing reads like "look how smart I am".
1. Extraneous details about satellites not being overhead at night -- as if government spy satellites operate on the same schedule with the same priorities as publicly known satellites.
2. Mentions things like them not scrubbing the EXIF data was a "mistake". How could it be a mistake if this was public information?
3. Says at the end "(this data is all public now)". Making it sound like he made the data public through his top notch sleuthing. It should read "this data was all public a week prior to my analysis".
If I were a betting man I would assume he started this exercise with the final assumed location in mind, and worked backwards from there to show how it could have been done.
I extended many of the assumed parallel lines in the photo, and they all tend to converge just below and to the right of the nose of the plane. And using the known distance between two of the stars to estimate the degrees per pixel, I get a margin of error of about 4 degrees vertically estimating where the horizon is.
With the casual change from "34-35 degrees" to "lets draw a line at 34 degrees" followed by the "and this is where the press release was" I got the vibe that this was parallel construction. Still some good sleuthing, don't get me wrong, but still.
You're right, it wasn't too difficult to guess the location without the stars. The base did happen to be right in-between the 34 and 35 degree lines (34.6 lat), and you can estimate pretty well with only one line
You're right. Vandenberg, Kirtland, Canon, Altus, and Little Rock are all matches for the latitude and there is no text explaining them away. Even the "it must be California because Northrup" is both (a) debatable and (b) doesn't exclude Vandenberg.
Also goes to show you the danger of tunnel vision.
If they flew one to some air base outside DC for the roll out the author would be shooting from the hip and who knows what facility he'd have zero'd in on. Sure it might have been the same one because "the same facility as the B2" is an easy just-so story but you don't really know with any certainty.
I don't know if I entirely buy the line on "34-35 degrees off the horizon." Unless we knew the focal length of the lens it was taken with, and which camera it was taken from, you don't actually know that. A wide-angle lens is going to have a much larger field of view than, say, a 100mm macro. And different camera systems have different angle of view for the same focal length.
Wow, it's cool to see astrometry.net get put to use! I worked on astrometry.net back in undergrad and grad school. If you have questions about how it works, I can answer them.
How does it work? astrometry.net uses 4-star combinations to define codes, then indexes the codes on the celestial sphere. The particulars of each of these phases matter, but that's the basic idea.
I used to be in astronomy and I always thought astrometry.net was one of the coolest tools in the field. It feels about as close to magic as you can get.
Astrometry.net is damn handy. I don't use it too much these days for myself except for annotating images, but I do have some friends just getting into astronomy and it's blind solve is amazing to help them.
Plus when I show them plate solving their minds are blown. Hugely useful tool.
I feel like the entire thread should have led off with "This is in fact a publicly known hangar in Palmdale and this was known as public information as soon as the unveiling event occurred, and thus this is just an exercise to show how locating the spot could have been possible from just this photo as a technical demonstration."
Because I guarantee there are going to be a hundred clickbait articles by various 'news' sites now in the next few days about the MASSIVE SECURITY BREACH of the USAF and how US SECRETS have been EXPOSED so and and so forth. When nothing about this is actually the case.
People have been doing photo analysis for a long long time, and have gotten quite good at doing geolocation based on details in the background that the average viewer ignores. Spy agencies are especially good at it. There's many a spy movie that offers glimpses into this, but The Good Shepard comes to mind.
Even as a joking kind of thing, the opposite has been done. There's a famous exchange between Neil DeGrasse Tyson and James Cameron about how the sky in the movie Titanic was wrong. Cameron accepted the challenge, and on a remastered version had the sky replaced to be historically accurate.
It's a game to some people. It's a challenge. It's self rewarding in this case, sometimes there might be actual rewards of sorts. Hell, I've spent more hours than I'd like to admit on Nasa's website where you use the images taken by the astronauts on the space station and align them on the globe. I got nothing out of it other than being entertained by the challenge, learned about some new places on the globe when I would try to find out why that place was interesting to the astronaut to snap that particular image, etc.
It's pretty clearly an exercise in astronavigation.
The livestream event itself mentioned it was taking place at the Northrop facility in Palmdale.
This author is not responsible for what clickbait farms do.
Aside: I'm not even sure this plane will end up doing flight testing somewhere secret in Nevada. They may just do it out of Edwards South Base, which is an "interesting" location not many in the public know about.
This trick happened for real with the HWNDU stunt back in 2017. 4channers found the flag based on the stars in the livestream, along with matching aircraft contrails to flight radar maps.
I feel like when you diss journalists, it should generally be for something they actually did wrong, not preemptively based on your imagination.
If they are so terrible, there should be no need to invent stuff.
When people read something like this post, and they are predisposed to the idea, it’ll reinforce their skepticism of „the mainstream media“. If you want to test yourself, make a bet of how many media outlets will run with the story in the manner outlined above, then check in a few days. My prediction: you won’t see it in the NYT, WSJ, BBC, or on CNN.
It is very interesting--and newsworthy--how difficult it is to be anonymous on the internet. Any person, place, or thing can be "doxxed" by seemingly innocuous background details.
It's worth noting that stars can be seen even in daylight with the right software.
That's because, while no individual star is visible, the exact angles between all stars is fixed, so you can do a brute force search of all possible orientations of the sky to find a matching one. In a 6 megapixel image of half blue sky, you effectively get an oversampling ratio of 3 million:1, so even with very bright sunlight obscuring the stars to the naked eye, your algorithm will pick them out.
This has only been bested by the 4chan takedown of the Shia LaBeouf's anti-trump flag, which was put up in a random location with a live stream.
"On March 8, the artists, abandoning the idea of a public webcam, raised a white flag, emblazoned with the words “He Will Not Divide Us,” in an “unknown location.” The livestream showed only the flag and the expanse of sky behind it. 4Chan and 8Chan (the forum where discussions that are banned from 4Chan go) snapped into action.... They used the star patterns visible behind the flag at night and the paths of planes flying overhead to confirm the location. A troll who lived nearby drove around honking until the noise was audible on the livestream. On the night of March 10, a group raided the site, took down the flag, and replaced it with a Pepe the Frog T-shirt and a “Make America Great Again” hat. The stream soon went dark again."
This is very cool sleuthing. Really enjoyed the walkthrough.
Though, if you would have asked me, I would have guessed Lancaster/Palmdale, because I know (a) That's where the unveiling was (b) I have seen B-1's fly over me while driving in Palmdale and (c) I know Edwards and a bunch of other spacey stuff is there. I feel like Captain Kirk looking for whales and saying "I think we'll find what we're looking for at the Cetacean Institute in Sausalito" - a far less methodical or impressive approach.
I mean, they livestreamed the rollout on YouTube on the Edward’s AFB YouTube channel and said it was at NG’s Palmdale facility (paraphrasing). I’m not sure any of this was a secret…
Which I think is also why the thread is so X-marks-the-spot / parallel-reconstruction-y. I think it's more of a "here's the kind of thing we can do with these star-matching tools, applied to an interesting image as an example" than an expose.
In fact, knowing the moment photo was taken from EXIF, there was no need to "plot the line". They could point the exact location by using orientation of the stars - stellar time (assuming the horizon is level which, in the airfield, is a very very good assumption to make).
Procedure is as follows:
Right ascension of the culminating star ("celestial longitude", taken from the star catalog/star map, of any star which is right above the celestial pole/"Polar star"), is the stellar time.
Knowing date (from EXIF), you get local solar time (which is equal to the stellar time on the spring solstice day, then makes a 24 hour loop over the year).
Difference between local time found this way and UTC time (taken from EXIF), gives longitude, because the Earth is round.
If EXIF does not contain the time zone, that gives 3 points in the U.S., one for each time zone, but it must be super easy to pick one. But most probably, EXIF contains the time zone.
The flock and skysat orbits he showed are for san francisco based planet labs satellites, the vast bulk of which are in sun synchronous orbits to always take photos in daylight. The screenshots amount to "it's dark at 1:30am."
Hahaha this is a neat bit of investigation but my favorite part is that the final conclusion is that the sexy new bomber is hiding……… at the Air Force base where they held the press conference announcing the new bomber!
One of my favorite internet happenings is when 4Chan (I think it was 4Chan) used stars and airplane data to get Shia LaBeouf's flag location and steal it. Not because of the political message, but because it was just entertaining and impressive in general. I think it ended up being in Tennessee somewhere. In any case, this wasn't exactly a secret, journalists were invited (though they had their cellphones confiscated at the gate, apparently.) If they really wanted, they could use obfuscation to make the sky, etc unrecognizable or in a totally different location.
Heh. This makes me remember how the same technique was used in the "he will not divide us" trolling campaign to locate a video feed of a flag. Internet Historian has a video about IIRC.
He seems to use the hour estimate of 1:30am being a time when very few satellites are likely overhead as a sort of confirming factor. (I guess because many satellites are sun-synchronous around noon, and the USAF would purposely time the photo to not expose the aircraft to observation).
But is that a very strong confirming factor? Is it likely that NG / USAF would be actively avoiding the potential for satellite imagery? Does it mean all the time it was being built, they avoided having it outside during the daytime?
[+] [-] gmiller123456|3 years ago|reply
Polaris is about .8 degrees away from the celestial North pole, so using it alone can give a reading off by .8 degrees, and the orientation in the photo is actually nearing the worst case.
The horizon doesn't appear in the photo. I have done celestial navigation strictly as a hobby for fun, and using a sextant without a visible horizon results in a few degrees of error at best. And that's with the benefit of the sensation of gravity helping you estimate. It is theoretically possible to estimate the horizon using the man made parallel horizontal lines in the image, but the author doesn't give any indication he did that.
Additionally, the author used the angle measurement tool in Stellarium to measure the distance to another ambiguous horizon. This is incredibly odd, as just clicking on Polaris in Stellarium will tell you its elevation. But even more odd, is he would have had to explicitly set his latitude in the app in order to get that screenshot.
[+] [-] whywhywouldyou|3 years ago|reply
1. Extraneous details about satellites not being overhead at night -- as if government spy satellites operate on the same schedule with the same priorities as publicly known satellites.
2. Mentions things like them not scrubbing the EXIF data was a "mistake". How could it be a mistake if this was public information?
3. Says at the end "(this data is all public now)". Making it sound like he made the data public through his top notch sleuthing. It should read "this data was all public a week prior to my analysis".
[+] [-] jaxomlotus|3 years ago|reply
[+] [-] gmiller123456|3 years ago|reply
[+] [-] microsoftdoes|3 years ago|reply
[+] [-] rootusrootus|3 years ago|reply
[+] [-] johnmcelhone|3 years ago|reply
[+] [-] lilyball|3 years ago|reply
[+] [-] cjensen|3 years ago|reply
[+] [-] dsfyu404ed|3 years ago|reply
If they flew one to some air base outside DC for the roll out the author would be shooting from the hip and who knows what facility he'd have zero'd in on. Sure it might have been the same one because "the same facility as the B2" is an easy just-so story but you don't really know with any certainty.
[+] [-] lambdasquirrel|3 years ago|reply
[+] [-] mierle|3 years ago|reply
How does it work? astrometry.net uses 4-star combinations to define codes, then indexes the codes on the celestial sphere. The particulars of each of these phases matter, but that's the basic idea.
[+] [-] dakr|3 years ago|reply
Fast forwarding a few years, we started using it at SOFIA Observatory to help speed up telescope acquisitions.
[+] [-] tetris11|3 years ago|reply
[+] [-] antognini|3 years ago|reply
[+] [-] malfist|3 years ago|reply
Plus when I show them plate solving their minds are blown. Hugely useful tool.
[+] [-] pash|3 years ago|reply
0. https://www.globenewswire.com/news-release/2022/09/20/251965...
[+] [-] wutwutwutwut|3 years ago|reply
[+] [-] mk_stjames|3 years ago|reply
Because I guarantee there are going to be a hundred clickbait articles by various 'news' sites now in the next few days about the MASSIVE SECURITY BREACH of the USAF and how US SECRETS have been EXPOSED so and and so forth. When nothing about this is actually the case.
[+] [-] dylan604|3 years ago|reply
Even as a joking kind of thing, the opposite has been done. There's a famous exchange between Neil DeGrasse Tyson and James Cameron about how the sky in the movie Titanic was wrong. Cameron accepted the challenge, and on a remastered version had the sky replaced to be historically accurate.
It's a game to some people. It's a challenge. It's self rewarding in this case, sometimes there might be actual rewards of sorts. Hell, I've spent more hours than I'd like to admit on Nasa's website where you use the images taken by the astronauts on the space station and align them on the globe. I got nothing out of it other than being entertained by the challenge, learned about some new places on the globe when I would try to find out why that place was interesting to the astronaut to snap that particular image, etc.
[+] [-] runjake|3 years ago|reply
The livestream event itself mentioned it was taking place at the Northrop facility in Palmdale.
This author is not responsible for what clickbait farms do.
Aside: I'm not even sure this plane will end up doing flight testing somewhere secret in Nevada. They may just do it out of Edwards South Base, which is an "interesting" location not many in the public know about.
[+] [-] johnmcelhone|3 years ago|reply
[+] [-] leoh|3 years ago|reply
[+] [-] _dain_|3 years ago|reply
https://youtu.be/vw9zyxm860Q
[+] [-] IfOnlyYouKnew|3 years ago|reply
If they are so terrible, there should be no need to invent stuff.
When people read something like this post, and they are predisposed to the idea, it’ll reinforce their skepticism of „the mainstream media“. If you want to test yourself, make a bet of how many media outlets will run with the story in the manner outlined above, then check in a few days. My prediction: you won’t see it in the NYT, WSJ, BBC, or on CNN.
[+] [-] godels_theorem|3 years ago|reply
It is very interesting--and newsworthy--how difficult it is to be anonymous on the internet. Any person, place, or thing can be "doxxed" by seemingly innocuous background details.
[+] [-] pifm_guy|3 years ago|reply
That's because, while no individual star is visible, the exact angles between all stars is fixed, so you can do a brute force search of all possible orientations of the sky to find a matching one. In a 6 megapixel image of half blue sky, you effectively get an oversampling ratio of 3 million:1, so even with very bright sunlight obscuring the stars to the naked eye, your algorithm will pick them out.
[+] [-] folli|3 years ago|reply
[+] [-] skykooler|3 years ago|reply
[+] [-] sva_|3 years ago|reply
[+] [-] calvinmorrison|3 years ago|reply
"On March 8, the artists, abandoning the idea of a public webcam, raised a white flag, emblazoned with the words “He Will Not Divide Us,” in an “unknown location.” The livestream showed only the flag and the expanse of sky behind it. 4Chan and 8Chan (the forum where discussions that are banned from 4Chan go) snapped into action.... They used the star patterns visible behind the flag at night and the paths of planes flying overhead to confirm the location. A troll who lived nearby drove around honking until the noise was audible on the livestream. On the night of March 10, a group raided the site, took down the flag, and replaced it with a Pepe the Frog T-shirt and a “Make America Great Again” hat. The stream soon went dark again."
[+] [-] amiga386|3 years ago|reply
[+] [-] daveslash|3 years ago|reply
Though, if you would have asked me, I would have guessed Lancaster/Palmdale, because I know (a) That's where the unveiling was (b) I have seen B-1's fly over me while driving in Palmdale and (c) I know Edwards and a bunch of other spacey stuff is there. I feel like Captain Kirk looking for whales and saying "I think we'll find what we're looking for at the Cetacean Institute in Sausalito" - a far less methodical or impressive approach.
[+] [-] vba616|3 years ago|reply
[+] [-] mrexroad|3 years ago|reply
[+] [-] rob74|3 years ago|reply
[+] [-] cf100clunk|3 years ago|reply
[+] [-] rognjen|3 years ago|reply
[+] [-] paxys|3 years ago|reply
[+] [-] darknavi|3 years ago|reply
[+] [-] boardwaalk|3 years ago|reply
[+] [-] bri3d|3 years ago|reply
Which I think is also why the thread is so X-marks-the-spot / parallel-reconstruction-y. I think it's more of a "here's the kind of thing we can do with these star-matching tools, applied to an interesting image as an example" than an expose.
[+] [-] anovikov|3 years ago|reply
Procedure is as follows:
Right ascension of the culminating star ("celestial longitude", taken from the star catalog/star map, of any star which is right above the celestial pole/"Polar star"), is the stellar time.
Knowing date (from EXIF), you get local solar time (which is equal to the stellar time on the spring solstice day, then makes a 24 hour loop over the year).
Difference between local time found this way and UTC time (taken from EXIF), gives longitude, because the Earth is round.
If EXIF does not contain the time zone, that gives 3 points in the U.S., one for each time zone, but it must be super easy to pick one. But most probably, EXIF contains the time zone.
[+] [-] bunabhucan|3 years ago|reply
[+] [-] pbronez|3 years ago|reply
[+] [-] kevmoo1|3 years ago|reply
[+] [-] partiallypro|3 years ago|reply
[+] [-] mpsprd|3 years ago|reply
[+] [-] supernova87a|3 years ago|reply
But is that a very strong confirming factor? Is it likely that NG / USAF would be actively avoiding the potential for satellite imagery? Does it mean all the time it was being built, they avoided having it outside during the daytime?
[+] [-] supergirl|3 years ago|reply
* he got the approximate latitude from the stars (34 or 35)
* independently, he found the most plausible airbase based on other information and then noticed that the base is indeed around that latitude