Inspired by the NY Times work using 3D models, I proposed to Documented NYC creating a 3D model for their climate change reporting, specifically for a story about a street in New York City that experiences frequent flooding. Unfortunately, due to web developing issues, the 3D model couldn’t be published when the story ran in April 2023. However, the process taught me valuable lessons about optimization and workflow that I've used in later projects.

I used photogrammetry, a technique that creates 3D models from still images. I scanned 183rd street, between 90th and 91st avenues, in Jamaica, Queens, a 270-foot long section lined with two-story homes.

For the capture, I used three DSLR cameras on a rig similar to this one from NY Times R&D. I'd used a similar rig on previous photogrammetry projects and found it extremely helpful.

I captured 5,272 images in raw format, which I processed in Photoshop before sending them to Reality Capture for reconstruction.

Lessons learned

The entire process was very time-consuming. Processing 5,000+ raw images and converting them to JPG takes time. I also had to manually add control points in areas where Reality Capture struggled with alignment. But by far, the most complex and time-consuming part was optimization and simplification. The street's characteristics didn't help. Cars, fences, gates, and light poles all added significant complexity to the scene. Since this model was meant for web embedding, optimization was critical and required advanced knowledge of Maya or Blender.

I chose a cloudy day for the shoot, which helped considering the entire capture took about three hours. During the capture, some parked cars were moved (about 4 times). Each time this happened, I took notes and returned to photograph the updated scene. When processing, I had to remove photos that included cars that had moved. The field notes were extremely helpful.