Hello, a quick update to the solution I did, while not a perfect solution if one is changing zoom, it might be of use to someone else to know the slightly odd means that I used, with a few tweaks you can even tweak it to find each edge.
The camera region its self was a 1x1 image, with four image points representing its edges, it also contained information for how much to offset the camera by on the Y axis(one can also add the X axis to the same concept).
Then each tick, from the players position(Not the invisible object representing the camera) I checked a point using a system statement exactly 128 pixels under the player(used 128 as it seemed to cause the least amount of camera snapping as I lerp the camera), if the statement picks more then one object in the particular group, it will use its top image point-Y offset to position the camera, otherwise use the players X and Y.
If this confuses anyone I will see if I can clean it up, likely a bit rube Goldberg for most cases.