Nukunukoo

If Samsung did make the sensor, then why did they lower the resolution from it’s almost 5 year old 28MP BSI sensor? Not that big a difference and do not justify a faster readout on an engineering point of view.

I mention almost 4 years ago that if Nikon/Canon were to go mirrorless, one option is the Translucent mirror route (which make all their lenses still compatible) there is a lot of merit in it (1) You can have a single all-green sensor on the reflected plane (and also acts as a full frame Phase Focus, which is easy because the sensor contains only a single colour pixel) coupled with a second red-blue sensor (that can also act as both phase and contrast focusing elements). The advantage is that it reduces not only 2x2 bayer wide noise and artifacts, but dramatically reduces high contrast false colours as well. The result is extremely close to a Foveon sensor minus all of the processing overhead. This is a boon to us Photoshop users when it comes to very precise edge-detect editing, or (2) Instead of all green, one can use all-white and increase the reflectivity of the mirror a bit and the result is that you have an image with significantly higher Dynamic Range and (3) you can place extra Phase-Detect sites on the sensor itself for superior and fast fousing (where the reflected PD is the coarse and the in-plane PD is the fine). Sony chose (3). It's interesting to see how this all plays out in the upcoming tests.

I mention almost 4 years ago that if Nikon/Canon were to go mirrorless, one option is the Translucent mirror route (which make all their lenses still compatible) there is a lot of merit in it (1) You can have a single all-green sensor on the reflected plane (and also acts as a full frame Phase Focus, which is easy because the sensor contains only a single colour pixel) coupled with a second red-blue sensor (that can also act as both phase and contrast focusing elements). The advantage is that it reduces not only 2x2 bayer wide noise and artifacts, but dramatically reduces high contrast false colours as well. The result is extremely close to a Foveon sensor minus all of the processing overhead. This is a boon to us Photoshop users when it comes to very precise edge-detect editing, or (2) Instead of all green, one can use all-white and increase the reflectivity of the mirror a bit and the result is that you have an image with significantly higher Dynamic Range and (3) you can place extra Phase-Detect sites on the sensor itself for superior and fast fousing (where the reflected PD is the coarse and the in-plane PD is the fine). Sony chose (3). It's interesting to see how this all plays out in the upcoming tests.

Perhaps if Nikon can update the firmware so that the camera can read 4.2K-4.4K internally then downsample to 4K, the crop would not be as severe. Being 8-bit anyways would make it not difficult a task for the Expeed 5. Of course the possible side effect would be a shorter recording time since the process can generate more heat than usual...