im making a platformer in 2d. Level 1 is an introduction walkthrough of town, im making one large detailed image of the entire level that is 20000x2048 in size. ill go in and separate the layers (background, middle, foreground) and essentially just cut the image up in squares and stream them into memory as the player reaches them and add colliders to the image to act as surfaces, walls, etc… Is this an unorthodox / impractical method for level design or is it OK in terms of look, performance and workflow?
Also, what is a good DPI to create the image in for best quality/performance mix? Right now I used 72dpi.
We’re using something similar. The wall colliders in our project have debug sprites that are normally disabled, but are turned on to show where the collider actually is when graphics aren’t yet available. We’re using large images of the scenes with the colliders’ debug sprites activated to draw scenery and terrain around, since the game isn’t tiled.
I don’t think you necessary need to just slice it up, but it might depend on the level of detail of your artwork. Our approach is to take individual “objects” from the generated scenery and add them as assets so that they can be reused if needed in addition to being reassembled in their original scenes. Slicing is only done for large pieces of terrain.
2 Likes
I was thinking this route would be the best since there will be parallax movement on layers I can just create assets of trees, clouds, etc. This way they can be reused and move independent of each other.
Thanks