The fourth workshop on Image Sensors for Precision Astronomy addresses the challenges of making state-of-the-art measurements with imperfect sensors. This article is associated with the opening talk at this workshop. The article includes a few key examples of sensor distortion with their possible impacts on today’s science goals. It also presents a few recent developments to mitigate their effects.

Tearing patterns affecting flat-field frames in charge-coupled devices are a visually striking obstacle to performing pixel response nonuniformity (PRNU) corrections. These patterns can be explained by lateral field distortions, caused by the nonuniform distribution of holes in the channel stops between sensor columns. Over the course of Large Synoptic Survey Telescope camera development, a number of practical fixes have been suggested to get rid of tearing. However, applying these fixes to our 16-channel Teledyne-e2v sensors leaves at best a distortion pattern at the vertical edges of every segment. Our working hypothesis is that the origin of the tearing is the parallel clocking itself, which moves the holes that are present in the channel stops regions. The efficiency of these transfers depends strongly on the details of the clocking operations, resulting in the observed variety of distortion patterns. Removal of most of the distortion patterns can, therefore, be achieved by executing a purge operation, which flattens back the hole distribution in the channel stops, immediately before acquiring a frame. A more effective solution is to switch all clocking operations to use a bipolar voltage set.