"He was like 'Name any mountain on Earth,' and I was like, 'I don't know, Mount Everest.' So he goes on Wikipedia, gets the latitude and longitude coordinates... and in about 28 seconds, delivered a 3D model of Mount Everest and all the surrounding mountains in that grid from the data. He's like, 'If you give me a couple of days we can take it for a ride...'"
It was a revelation for a team that, up to that point, had been struggling to use procedurally generated 3D noise fields to partially automate the process of creating convincing-looking mountains, with largely unsatisfying results. "We were doing a lot of research of real-life mountains and the mountains themselves we were generating just didn't have any of the kind of personality and unique features of some of the craziest mountains we were seeing that actually exist in the world," Batty told Ars.
But with the ASTER data providing a free, complete, easy-to-convert map of the best snowboarding runs nature had to offer, all that work seemed beside the point. "We thought, 'Wow, are we done?'" Batty recalls.
