Effiency question: How to improve on large images as game map

:information_source: Attention Topic was automatically imported from the old Question2Answer platform.
:bust_in_silhouette: Asked By jevans

Hi, I’m very new to Godot (and game making in general) and I suspect my current system is inefficient and want to correct before it becomes an issue. I’m making a top-down, 2d game. So far, I’ve made the maps out of large images (usually one large picture for the map, with a mostly-transparent copy with just the tops of doorframes and other overlaps isolated layered over it and the player). (The images are about 16000px wide at the moment). This does work and I like how it looks, but I’m already starting to see performance issues. From what I’ve read about efficiency, it sounds like this much transparency is probably causing issues. I suspect the giant map image file is too, but I’m not sure the best way forward. Would it be better to break it into smaller sprites that add up to the same total? I’ve thought about switching to tilemaps for Godot’s optimization, but most of the planned maps are large, non-repeating pictures. Would converting them to tilemaps be worth it, even if most tiles don’t get reused? I had also planned to fake 3d spaces by turning colliders and sprite visibility on and off depending on the player’s direction of approach. If I switch to tilemaps, would it be worth it to make ‘upstairs’ sprites into tilemaps too?
I know this is probably really beginner stuff, so thank you for any help you can provide.