Not like most telephones launched as of late, Apple’s new iPhone SE has only one 12-megapixel rear digicam lens. However it seems that one digicam remains to be fairly good, not less than in well-lit conditions, my colleague Dieter Bohn present in his iPhone SE overview.
For portrait photographs, the iPhone SE’s digicam makes use of machine studying to estimate depth of subject, and Ben Sandofsky, one of many builders of cell pictures app Halide, took a closer look at how at how portrait photographs taken by the SE’s single digicam truly work.
The important thing characteristic of the iPhone SE’s portrait mode is that may do one thing referred to as “monocular depth estimation,” which is enabled by the iPhone SE’s A13 Bionic processor (the identical processor within the iPhone 11 and 11 Professional). That processor and the depth estimation permits the cellphone seize depth maps for photographs in another way than, say, the iPhone XR, which additionally has a single lens. Sandofsky discovered that the SE may even estimate depth for flat photographs.
In a single instance, Sandofsky took an image of this picture with an iPhone XR and an iPhone SE:
On this picture slider, you may see how the 2 telephones estimated depth. The iPhone XR’s depth map is on the left, and the iPhone SE’s is on the fitting.
The iPhone SE’s depth map isn’t an ideal illustration of the particular depth of the scene by any means, however fairly darn cool that the SE can estimate the precise depth of a flat picture fully utilizing machine studying.
I extremely advocate taking a couple of minutes to learn Sandofsky’s blog in full. He goes into much more element about how the SE’s digicam captures depth, and has a bunch extra examples of its depth mapping in motion — together with one instance the place it finds depth in a 50-year outdated slide movie of his grandmother!
And take a look at Dieter’s overview if you wish to see extra of his ideas in regards to the SE’s digicam.