Originally posted by DSims Remember that the pixel size on an APS-C 16MP sensor is almost identical to that on a 36MP FF sensor, so all other things being the same, what would be the difference?
The difference would be that, given the same field of view, the 36MP FF-sensor uses more pixels to resolve the same amount of detail.
For example, say you take the same shot at 50mm on the D800 and at 32,9mm on the K-5 and there is fine structure in the frame that covers 10% of the hight and 10% of the width of the frame.
The resolution of that crop will be 736x491 pixels on the D800 and 493x326 pixels on the K-5, now moiré patterns only form if the size of the parallel lines of the fine structure in the frame and the parallel lines that are formed by the rows and columns of pixels on the sensor are of similar size. So the fine structure has to be smaller, to produce moiré patterns on the D800s sensor than it would have to be to form such patterns on the K-5s sensor, therefore moiré is less of a problem for the D800 than it is for the K-5.
There are non-bayer patterns like the one the X-Pro1 uses, that produce less moiré, but they usualy need more photocites to produce the same MP-count.
The bayer pattern is more efficient but it needs an AA-filter, at least at lower MP-counts.
Keeping the sensor size (APS-C), the individual photocites have to be smaller (and thus capture less light) but larger microlenses in front of those photocites can make up part of that difference.
I'm thinking that if - even with a 36MP sensor - Nikon decided to make two versions with and without AA-filter, it's unlikely that pentax would use a 16MP bayer-sensor without an AA-filter.
But if they're using a new 16MP sensor anyway - and not the one ued in the K-5/30/01 - then it might be the same 16MP non-bayer sensor the X-Pro1 uses, which just happens to work without an AA-Filter.
Given how the Fuji X-Pro1 performs compared to the K-5 it would be a step up in IQ and certainly better than the 24MP Sony sensor used in the Nikon D3200.
-Jan