what if there was a way to “increase frame rate” by adding in some type of logic checker between two generated images? kinda like a comparison between two generated frames that would lead to more generated images that mimic movement? so like a filler between frames that would predict how something got to one shape to another using a set of properties that a generated object has, those properties could be weight, speed, gravity etc etc, it just depends on what object it is conceptualizing or constructing