The World’s First U2 Spy Plane Photo Shoot at the Edge of Space

> So many comments here are casual dismissals of the photographer’s style, and while I generally prefer a lighter touch in post-processing, there are some incredibly cool photos in here.

I kind of feel bad of the photographer, but I think his style has is probably collateral damage of generative "AI." I literally did not believe his photos were real, because the uncanniness and the subject matter just screamed AI-generated to me. Especially that photo of the pilots in the space suits. A lot of the night shots of the plane also looked like renders.

It's almost like it's too easy to fake cool, which then devalues cool, so to be believable someone an artist almost has to miss the mark in just the right way.

There's probably a deeper point there, but I can't make it.

I would not have guessed that this is a particular post-processing style (I no nothing about photography); my first impression was that some of the pictures were either AI generated (the general aesthetic is quite common in AI generated pictures), or 3D renders.

Are there good explanations for why AI generated pictures tend to emulate this particular post-processing style? Is it a simple matter of choice of training data, or does it have to do with the "unnatural smoothness patterns" of the generated (or in the article's case, post-processed) image?

Apart from that, I find it quite surprising that this is the first photoshoot. Isn't U2, first and foremost, a high altitude photography platform? Sure, they probably mostly cared about photographing the surface of the planet, and perhaps the plane's (fixed) equipment was even such that it could only photograph the surface. But did no one pilot, co-pilot, etc. ever care to try and take pictures with even a humble consumer camera in a U2?

Finally: could the post-processing style have been chosen on purpose to hide details which are (for whatever reason) still "sensitive information"?

The couple of photos by Blair Bunting in that article that have people in them look so over processed that I hesitate to believe they aren't entirely AI or otherwise computer generated.

While I'll stipulate that every artist is allowed to have their own style, I too would love to have seen some of these same images much less stylized. I want to see the plane! So I would not be buying a coffee table book with this overly stylized imagery, but I'll restrain myself from calling it shitty.

Part of the article stipulates that much of the plane is still classified to a certain extent, it could be possible that there were requirements to not show too much detail. That would lend to the highly stylized look to being used. I'd have assumed some photoshopping to paint over areas that needed redacting.

I thought the extreme HDR look went out of fashion when it became too easy for phones to do, and was overused.

These pictures don't have bad HDR artifacts (EG haloing) but even a casual observer will notice they don't feel natural at all. It's a shame because the framing and colors of those pictures are really nice, but the contrast (both high and low freq) gives them a completely synthetic feel. I'm guessing the sky was swapped out with a very dramatic overcast version, which also adds to the uncanny valley effect.

According to the article, the "colors ... are, importantly, not added for dramatic effect — these photos and all aspects of them, are real."

What processing are you referring to?

I honestly thought they were screenshots from Microsoft Flight Simulator and that this whole thing was going to end with him at a PC and an advertisement to join Xbox Live so you can flight the U2 as well.

I thought I was reading some fantasy article with CGI "photographs" littered all over. Yeah, it's bad.

I would say the F-22 and F-35 are _far_ more advanced than either the SR-71 or the U-2. I think spy planes have been made less critical and more vulnerable with advancements in spy satellites, drones, and air defense and so are not a focus of R&D.

“Complex systems won’t survive the competency crisis”

The X-37B seems like a likely candidate.

It's like asking why rotary phones design stagnated since the 60s, we just don't have the use case anymore, we have better tools for the job

IIRC The SR71 stopped being useful as a spy plane because missile defence improved to the point they could be shot down, and spy satellites took over the duties of reconnaissance.

These old airframes do get incremental upgrades to software/avionics/engines etc, and there is actually a SR-72 in development as a UAV.

Presumably:

1. Satellites can now conduct imagery intelligence at a similar or higher quality without alerting targets, conducting refueling, or risking pilots.

2. There are more satellites with higher maneuverability than before.

3. Unmanned high altitude drones will replace the existing U-2S role in the next decade or so; or already have. Consider the U-2S role in the Chinese balloon incident. The military wanted intelligence on the balloon that would not otherwise come from a satellite and therefore utilized the U-2S.

Honestly, I think so. I think reliance on computers did it.

I would say absence of evidence is not evidence of absence in this context.

What makes the U2 quite amazing is not just that it's a 70 year old design still in operation.

It's also that the U2 was developed by Lockheed Skunk Works on a $22.5 million contract, where they delivered 20 working aircraft in less than 2 years from project start and went $3.5 million under budget. They paid back the money. And as a bonus deliverable, the team also developed a new testing site with extra long paved runway, hangars, workshop etc. on a salt flat known as Area 51.

It's more that this is close to the best design possible for this mission, it would take decades and lots of money to develop another aircraft that does the same thing that might not produce that much of an advantage. There are lots of designs that have been around for ages because they are effective and there is no reason to replace them. Even in the civilian world, there are lots of aircraft types that have been around for 50+ years because there is no real reason to get rid of them as long as they can be upgraded to the latest systems.

It's really only when an aircraft can no longer be upgraded or its mission no longer exists and can't be adapted easily for another one is when it's retired.

Those aircraft were designed to maximally exploit the physical properties of statosphere via airfoils, gas turbines and materials science. There's no more there there other than more efficient turbines and materials which aren't really needed to accomplish the limited mission. Anything higher or faster would be less an aircraft and more a rocket so much more logical to jump the mesosphere and go straight to space.

To echo other commenters, it's simply because it's not needed as heavily anymore, and it's a very optimal design for the purpose.

Consider what satellite imagery was like six decades ago. The CORONA program used massive and complex satellites to take photos on film reels, then jettison and recover them using parachutes and catch them on a hook from an aircraft. Yes, really. Back then, it was much more effective to send a plane over and take some images manually.

Very soon after, even as early as the mid 70s, digital image processing and transmission made these obsolete with the launch of the KH-11 series. By this point, it was substantially faster, cheaper, and more reliable to use satellite imagery rather than spyplanes for general purpose surveillance.

The other thing to consider is the rise of surface to air missiles and air defense systems. These greatly changed the calculus of manned surveillance flights.

https://en.wikipedia.org/wiki/CORONA_(satellite)

https://en.wikipedia.org/wiki/KH-11_KENNEN

https://en.wikipedia.org/wiki/1960_U-2_incident

The surveillance and EW electronics on the U-2 are kept up to date. You'd have to think long and hard about the value in developing an all-new successor to the airframe, rather than an incremental update to the U-2. The design is already very well suited to a very specific flight envelope. (as a practical matter, an all-new replacement made today targeting the "piloted and subsonic and fly as high as possible for a long time" would be required to have working stealth features and would be huge and cost a zillion dollars, and nobody involved wants to pay for that)

Regarding the aircraft comparison: the U-2 is surprisingly cheap to operate. It's brilliant. The SR-71 pretty much pegged the meter in terms of operating cost.

Agreed, I can't tell if it is real or all generated?

No kidding - despite the fact that the images are all credited as "Photo by", I simply don't believe they are photos. They look more like videogame loading screen artwork (especially the images with engineers working on planes)

I thought it was universally known that photographs usually have post processing effects (filters) applied to them. Post processing effects allow the artist to place their hand on the picture. It's clear that the photos have been touched by the artist and they have achieved a look that the photographer was going for.

Most photographers alter the raw photograph data.

All of his photos are highly edited https://www.blairbunting.com/Portfolios/thumbs

I don't get how you go and get access to something that not many people get to see, and then you edit the photos so much that they look computer generated.

Looking at his portfolio it seems like he his better at photoshop than at taking photos.

The article takes a lot of creative liberties to make this article, looks like.

I launched a helium balloon to ~90000' with a couple of GoPros. The footage looks similar.

> If you extrapolate, the earth would be much smaller in these photos.

At that height, the horizon would appear nearly as a perfect straight line. With a pinhole camera, it would be projected as a straight line in the image plane. Thus, the curvature here has to do with the wide-angle geometry of the lens. If the camera was pointing higher, you would see the horizon curved in the other direction (which would look even cooler!).

The curvature does look sort of uncanny but then again, we don't often see photos from this vantage point. While some of these are wide angle photos which can have lens distortion, photo processing tools also have lens correction options to remove such distorions. Looking at the photos with the aircraft's wings visible, it doesn't appear as though there's much lens distortion remaining.

One of the coolest aspects of the photos to me is the way you can see terrain in a fairly typical aircraft view underneath and still see the curvature of the horizon toward the edges.

Since it's on a photography-centric site, I was a little disappointed there was no discussion of the camera gear or lens selection. I was also hoping to see camera EXIF metadata alongside the images which would typically reflect lens type, focal length, shutter speed, aperture, etc.

Consistency and uniforms? Uniform dress is kind of a keystone for military service.

I assume someone connected has the supply contract and thought it would be a great idea if their clothes were worn at all times. Like why on earth are sailors dressed in blue camouflage. I think it also helps hide how out of shape many of them have gotten which would be more obvious if they still wore khakis.

Once upon a time, I used to support these very aircraft. I worked as a technician on the flightline. As an airman, you show up to work every day in a uniform. For us, it was either a set of camo BDUs or 'blues.'

No one I knew that I worked with preferred blues over BDUs. One would wear blues only if absolutely required.