You can scan a location for a pixel and use that data to trigger something else but you cannot detect pixel movement with expressions or scripting. What that means is that you can pick a point in your comp and look for say red pixels. If there are none then do one thing, but if a red pixel shows up then do something else.
You would have to use development tools to write your own plug-in. It would probably be quicker to insert some keyframes.
SImply keyframe it. It's way too complicated to do such stuff with expressions and would be a major performance hog. Pretty pointless to even think about it when something can be done with more conventional techniques.
I should really do it manually since there's no other way that's simpler than just keyframe.
My intention was to write the script, so I can put this into production pipeline to use it over and over again if this was possible.
Thank you !
Thanks for the comment !