Tilemap VS SpriteBatch VS Sprite List

[Croweley]

Hey all,

 

I'm about to use a tilemap for our game, which needs

 

1- A ground layer

2- A layer for players and non-tile objects (which are groups, not sprites)

3- An upper (above player) layer

 

I'm curious about the best method to go about this, performance-wise. My first thought was to maintain a list of tile Sprites, indexed from an Atlas (which would be my tileset). I'd have one list for ground tiles, and one for upper tiles. Then I could simply maintain 3 Groups (ground, players, upper). With the upper and lower groups utilizing SpriteBatch, would this perform better than Phaser's built-in tilemap system? (since I won't need to maintain 2 canvases for the tile layers, and would only process the tiles I need to render).

 

If there is a better way (ie- using tilemap) then I'd love to hear about it!

[Croweley]

Perhaps there is a better way to word this. I have 2 questions:

 

1- Which would perform better - two 1280px x 720px TileMapLayers, or two SpriteBatch groups containing 40px x 40px sprites to fill a 1280px x 720px resolution?

 

2- If a TileMapLayer has only one visible 40x40 tile in it, is its entire 1280 x 720 canvas rendered (ie- all pixels processed, or just the 40 x 40 tile)?