We’ve been getting fooled by photos of the night sky for years; now, tech is making it even easier.
By Emma Roth, a news writer who covers the streaming wars, consumer tech, crypto, social media, and much more. Previously, she was a writer and editor at MUO.
Mar 16, 2023, 9:30 PM GMT+6:30
Share this story
Venus beside a crescent Moon on July 15th, 2018.
Venus beside a crescent Moon on July 15th, 2018. Image: NASA / Bill Dunford
When astronomer Tyler Nordgren first got involved in astrophotography in the ’90s, he noticed something very off about the postcards, posters, and other photos he’d see when living and traveling in the American Southwest.
“One of the big things that struck me at that time was the number of pictures I’d see that show the buttes in Monument Valley with a full moon rising behind them,” Nordgren recalls. Nordgren had been to that exact location in Monument Valley, and he knew the Moon didn’t rise in the position shown in the photos. “And even if it did, the shadows on the Moon are utterly different from the shadows on the buttes.” Even in a time before widespread Photoshop use, it was clear that something was up: the photos were fake.
A dark, starry night sky is undeniably beautiful but also remarkably difficult to snap a decent photo of. This week, Samsung drew criticism for the technology its newer phones use to “enhance” photos of the Moon. A user on Reddit, ibreakphotos, conducted an experiment by creating a blurred photo of the Moon and then taking a picture of it using their Galaxy S23 Ultra. Even though the photo was completely blurry, their Samsung device appeared to add details to the image that weren’t there before, like craters and other marks, calling into question whether the highly detailed Moon photos people have been taking with their Galaxy devices really are photos of the Moon. While Samsung’s Moon fakery has ignited a debate around the appropriate way to photograph the Moon, the truth is that people have been faking the night sky for a long time — even without the help of artificial intelligence.
“This is something that in the astrophotography community comes up quite a bit,” Nordgren tells The Verge.