cameras for 2D layers are…..a bit “different” because you don’t use them. I mean, you Could just take your entire 2D composition and turn it into a 3D layer in a new comp. Then animate a camera around. That’s one way. But cameras are cumbersome, this would render slower, and add unnecessary complexity.
Instead. You use Professor Farnsworth’s method of space travel. Since he couldn’t move the space ship fast enough, he just moved the entire universe in the opposite direction.
So just like what you’re looking at there, the entire animation is done on a “sheet of paper” as they reveal in the end. You’re just moving that sheet of paper around in front of your viewport.
You can precomp, and do all your “camera moves” in the final composition. The annoyance here is you have to go back and forth since framing your animation is done in a different comp. Or you can parent things to nulls. The annoyance there is every time you parent, if you need an expression, you’re going to have to get world position rather than the object position. Not a big deal, but definitely annoying.
They are using wiggle, too heavily in my opinion. But I think the overall motion, they’re just animating by hand.