Doing some background research on stepper motors. If we go with zeotrope as Richard is suggestion, this may be less of an issue as we would need fewer steps, but otherwise if we assume we need 2000 images and 2000 unique masks, we have 2 problems (depending on how we tackle the issue I know… but let’s run with it for now):
How to get that many unique masks?
How to get a motor that can handle 2000 unique positions?
I looked into micro-stepping for a stepper motor. Most steppers have 200ish steps, and then you can micro-step between the major steps (8, 16, 32 micro steps per step). That gets us to 2000 easily, but the microsteps aren’t perfect - here’s a nice article showing the accuracy of microstepping - http://hackaday.com/2016/08/29/how-accurate-is-microstepping-really/ . If we go that route, this is something we’ll need to design around.