This photo was taken with a Google Pixel phone in almost complete darkness by Florian Kainz.
Though there is still a lot of work that needs to be done – a Google researcher may soon have an answer for easily capturing great photos at night. Smartphone cameras have come a long way since the days of flip phones but due to the constraints of size, lighting – they don’t always take the best pictures under certain conditions.
Florian Kainz, a software engineer on Google’s Daydream team, developed a technique that allowed him to take photos in near total dark conditions with the Google Pixel and Google Nexus 6p phones. His results are comparable to photos taken with a DSLR on a tripod. Amazing, huh?
Kainz accomplished this by using burst photography where many shots are taken of the same scene. The resulting photos are analyzed to remove sensor noise. This research is built on the foundation of another experimenta app, SeeInTheDark by Marc Levoy.
He built a typical Android camera app that would give him control over key parameters such as ISO, focus distance, and exposure. Instead of taking a single photo – it takes many shot in a short period of time. Single photos taken on a moon lit night were grainy. However, when Kainz computed the mean of all the images and accounted for movement, he was able to develop very clear photos.
“Arriving at the final images required a lot of careful post-processing on a desktop computer, and the procedure is too cumbersome for all but the most dedicated cellphone photographers.
However, with the right software a phone should be able to process the images internally, and if steps such as painting layer masks by hand can be eliminated, it might be possible to do point-and-shoot photography in very low light conditions,” Kainz wrote in a blog post.
Kainz still faces a few obstacles to make this technique available to the masses in an a one-click solution. This approach currently only works if the phone is resting on a solid/sturdy object or tripod.
While Kainz did most of the work by hand, it’s not very hard to imagine most of this process being automated and making its way to a widely-available camera app in the near future.
Check out an album of Kainz’s photos processed using this technique.