One of the major changes included in the Galaxy S7 and Galaxy S7 edge over their predecessors is the 12-megapixel camera that features dual pixel technology.



Samsung's previous flagship smartphones contained 16-megapixel camera modules, but with the Galaxy S7 this has changed as the South Korean company is now focusing on quality rather than the megapixel number.
Although many have been pretty skeptic about the performance of Galaxy S7's 12-megapixel camera, early tests show improved quality over the Galaxy S6 and Galaxy Note 5 cameras.
The most obvious is the blazing fast focus due to the dual pixel technology. As a matter fact, the Galaxy S7 camera might offer the fastest autofocus feature ever included in a smartphone.
BRITECELL camera technology not yet ready for prime time But what about the module camera? It looks like Samsung's BRITECELL camera technology isn't quite ready for prime time since the South Korean company has decided to include a Sony module inside the Galaxy S7/S7 edge.
The folks over at Chipworks have decided to open up the Galaxy S7 edge and see what's inside. One of the things they have managed to confirm is the fact that the 12-megapixel camera inside the smartphone has a Sony sensor.
They haven't been able to confirm the model number, but rumor has it that it's a Sony IMX260, the newest camera sensor coming from the Japanese company.
“Our lab staff have removed the imaging chip from the 12 MP camera module and have found a Sony back-illuminated (Exmor R) CMOS image sensor. We were expecting to find TSV arrays around the periphery of the active pixel array, corresponding to Sony’s stacked chip (Exmor RS) technology platform,” reads Chipworks' article.
Also, the 1.4-micron pixel size of the Sony sensor has been confirmed as well, along with the camera module measurements: 12.1 x 12.1 x 5.4mm.
On the other hand, the 5-megapixel front-facing camera includes a Samsung S5K4E6XP sensor that measures 1.34-micron pixel.
In regards to the dual pixel technology implemented in the Galaxy S7 edge, the folks at Chipworks are writing that “the full chip PDAF functionality, which requires dual readout from each pixel, was implemented with a multi-chip solution rather than a stacked (CIS + ISP) solution.”