I've made a Tiled game. Right now I'm stress testing my phone's capabilities by increasing amount of nodes in the scene. There's physics based stuff, AI movement, Day & Night system, particles popping out here & there & plenty of other stuff going on under the hood for my scenes. What I want to know is there a performance difference in using 16x16 tiles, what I have now, versus using 32x32 tiles? These tiles are basically just an image that was added to the scene; they don't have any physics bodies or anything else of that sort. They do have properties I set upon them when making the map in Tiled, but I don't think that has any performance impact. Each map has several layers (background, vegetation, spawn points, buildings, sometimes a few more). Here is a code snippet of how tiles are rendered for 1 such layer:
if([map propertiesForGid:tileGid][@"shrub"])
{
SKSpriteNode *tile = [layer tileAtCoord:coord];
tile.name = @"vegShrub";
[self addChild:tile];
}
else if([map propertiesForGid:tileGid][@"tree"])
{
SKNode *tile = [[Breakable alloc] initWithWhole:[atlas textureNamed:@"tree"] broken:[atlas textureNamed:@"tree-stump"]];
tile.position = [self pointForCoord:coord];
[self addChild:tile];
[layer removeTileAtCoord:coord];
}
If I use 32x32 tiles over my current 16x16 tiles, will I somehow free up some memory or "relieve the load" off the system?