How Google wants to improve the night on smartphone photos [Exclusive Guide]

1:16 am April 28, 20179105

if the technology for made huge progress, it is clear that our phones are far from being able to compete with professional cameras in terms of night photography. However, a Google engineer think changer thanks to an application which, thanks to its algorithm would be able to produce a stunning night photography.

google photo nuit

while he captured images of the Golden Gate Bridge with my digital camera, Florian Kainz, a Google engineer, received a challenge from one of my colleagues: resume exactly the same photo, but with a smartphone this time. A bet at first sight difficult in the light of current technology, as he points out in my blog search.

A Google engineer develops an application to improve photos of night

but the engineer agrees and sets in the development of an application on Android inspired by “SeeInTheDark”, a software app capable of producing pictures of night of good quality even when the light is low, by combining several shots . A principle similar to that of the HDR .

Then begins the work of Florian Kainz. After reviewing the operation of SeeInTheDark, the engineer is then developing my application. «He explains: “ I developed a simple app with a manual control over the exposure time, sensitivity to light and the focus distance “.»

When the button is pressed, the application is waiting for a few moments and takes then 64 pictures in a row. The result, once the images are processed, is so totally amazing. More grain: images are sharp and seem to have been taken by a very high-end camera (they were actually captured by a Google Pixel and a Nexus 6 p).

Of course, we imagine that this process will take a lot of space on the memory of our devices. Nonetheless that this application, for the moment at the stage of development, for a monumental potential. For the record, Florian Kainz won my bet and its images can be found in my album Google Photos .

The photograph of the Golden Gate Bridge at the origin of the challenge:

google photo nuit defi

the same photo taken by the engineer, and processed by the application:

google photo pixel nuit