Samsung's Latest Camera Sensors Promise Improved Low-Light Shots

Next Story

Alert Phone Lets Your Loved Ones Warn You When They've Wandered Off

You know what sucks? Taking pictures in a bar. You might be having the best time in the world at that bar, but the pictures just never do it justice. The bartender gave you a free drink, that cute girl has totally been staring you up all night, and your pub trivia team is destroying the competition. You take a snapshot on your phone to capture the moment — and it’s blurry, blueish garbage. Instead of capturing the fact that you and your friends are a bunch of lady-killer trivia masters, you look like freaky-deeky Smurf ghosts.

Samsung understands your pain — that’s why they’re working on two new smartphone camera modules which they claim ought to improve the quality of low light photos.

First up, the 8 megapixel S5K3H2 sensor. It’ll shoot 1080p video at 30 FPS, but if you’re willing to drop the quality down to 720p or VGA (640×480), it’ll shoot’em at 60 FPS or 120 FPS respectively. Oh, and those 120 FPS videos can be played back in slow motion — perfect for capturing every glorious “I told you to stop” millisecond when your friend is yacking up his lunch outside of the bar.

Next up: the 12 megapixel S5K3L1. It’ll do 1080p/720p/VGA videos at the same frame rates as the sensor mentioned above, but it also packs a specialized pixel-correction chip built to try to automatically clean up any image distortion.

The big features on both sensors, however, is the back-side illuminated technology within. For those who don’t speak mega-nerd, that just means that they’ve used some incredibly fancy engineering tricks to move some components around inside the camera in a way that allows considerably more light to hit the sensor. More light = better quality photos.

Samples of the modules should start reach manufacturers in about 6 months, which means we should start seeing this things in devices within a year or so. Creepy blurry Smurf Ghosts, be gone!

blog comments powered by Disqus