The Nikon D800 has been announced. It’s a camera with full sensor and video capabilities. Philip Bloom had the following comments: “I am disappointed by 50p/60 only being in 720p, that’s a lot more binning again…shame it’s not 1080p.”
And further down the page he says: “Now my biggest concern is the size of the sensor. Actually let me be clear, not the physical size but the megapixels. Now this is based up video functionality not stills. Yes, I am interested in stills especially for timelapse but the most important function for me is video. The D800 has a WHOPPING 36.3MP sensor giving us a resolution of 7360 x 4912 pixels which is just insanely large. In comparison the 5DmkII has a more conservative 22.1MP and the brand new top of the range D4 has, some may say inadequate, 16MP. Personally the lower the better for me. 16MP is MORE than adequate for stills unless you need monstrously large prints then the D800 is what you need. More megapixels for video causes us these issues: Image size needs to be drastically reduced to get down to the 1920×1080 full HD video image and to get there in the Canons has meant line skipping, information binning, which gives us moire and aliasing. Not nice. In fact very ugly. I am not sure how bad the binning is on the D800 but that is a lot of information to throw away. It is 1.6x larger than the 5Dmk2 so that worries me. I have noticed some image issues on the video below but never judge image quality based upon a compressed web video. I need to see the original, or better still, shoot with the camera to form a proper opinion.”
What is that “binning” he’s referring to? I didn’t know until I found this article. The article states: “Pixel binning is a clocking scheme used to combine the charge collected by several adjacent CCD pixels, and is designed to reduce noise and improve the signal-to-noise ratio and frame rate of digital cameras.” It explains in detail what binning is, and then concludes: “The primary benefit of pixel binning is to improve the signal-to-noise ratio in low light conditions at the expense of spatial resolution.”
Interesting, heh? It could explain why GoPro HD Hero2 cameras are noisy in low light conditions, if I understand it correctly, that is.
visualsproducer.wordpress.com 
Binning, effectively, is rescaling/resizing the image down. It’s quite tricky to do this on-sensor for RGGB type arrays for obvious reasons. Which I imagine is why skipping is used instead. If they manage to read the entire sensor fast enough then binning is as easy as rescaling the frame.
LikeLike