Here's the comment I was remembering: Rich Altmaier 3:23 PM (5 minutes ago) to me The sensors Intel was producing, now about 7 year ago, did not support captu

Here's the comment I was remembering: Rich Altmaier 3:23 PM (5 minutes ago) to me The sensors Intel was producing, now about 7 year ago, did not support capturing the full set of frame pixels at one time. Rather pixels could be copied into backing store from just one pixel row at one moment, then read out, then copy the next row and read out. So the single row of pixels was from the same moment in time, and subsequent rows had to spread over about 1/120 second of time. So all images were time delayed from top to bottom of the frame. This method, common to smart phones, is the reason fast moving objects appear bent. e.g. helicopter blades are bent in mobile video. Secondly I wanted an enhancement, not available at that time in any modern DSLR camera, to adjust the capture image brightness based on ambient illumination changes. This would prevent the typical gym lighting, fluorescent oscillating at 1/120s, from a beat frequency with the video shutter, showing slow moving bands of light/dark across the screen. Rich