Technology for low lighting imaging
High-Dynamic Range (HDR) is a technology more efficient for photo capturing than for video. To take an HDR photo, the sensor takes different shots of the same scene at different exposure times and then mixes the shots by selecting the best. Basically, an HDR picture is a mix of pictures that takes out the best exposure time of the specific areas of the pictures according to their illumination. A longer exposure time is used for low illuminated areas while a shorter one is intended for high illuminated areas. For this reason, HDR photos should normally be taken for static scenes and not for moving objects. If the object is moving between the shots, the HDR photo will show motion artefacts.
In the case of videos, this problem is more evident.
There are different techniques how to increase the Dynamic Range of an image sensor. ON Semiconductor applies the multi-exposure technique which, like for HDR photo, takes more shots of the same frame at different exposure times and then interpolates the shots creating a single HDR frame. In order to perform a good quality HDR video, the sensor needs to do some motion compensation processing of the video itself in order to avoid motion artefacts.