J Davis
Veteran
I'm speaking of a what if... what if the chip could be cropped in this new camera as a new feature. With pixels at a known size on the GH1 chip, what would be the image area covered by 1920x1080 pixels. And would that 1920x1080 pixel area be covered by the s16mm or r16mm gate sizes so that corresponding image circles would cover it.
In other words, to have the possible benefit of avoiding pixel binning, or whatever methodolgy they would use on the af-100 to downsample 12MP to 1920x1080, to avoid that in a possible cropped feature, what would that image area have to be for a 1:1 sensor pixel to recorded pixel relationship?
Right, okay, I see what you're saying. That isn't gonna work.
First of all, to answer your question, I think you're asking what the image size would be for a 1920x1080 central crop out of the 12mpix sensor, right? Answer is it'd be about the same as a 1/2" chip. So yes, super16 would easily cover that.
But, if they do this camera right, that won't be worth anything other than, at best, standard-def res. Because part and parcel of why the camera will work without aliasing is because it will have a properly-engineered optical low pass filter in place. That will reduce and blur the 12mpix source down to a managable, alias-free 1920x1080. And if you go sub-windowing out of that, you'll get maybe 400x240 of actual resolved data.
So ... no. Not likely gonna work.
What dadoboy (I think) is meaning to say is to have the cmos windowed area the the pixel dimensions it needs to be so that after debayer it is 1080p.
Funnily enough, the 2/3 spec'ed scarlet is just that.