LOD in an Online-Top-Down

Hey there,
I am curious how LOD works. I mean I have a basic understanding of how it works and using multiple meshes, etc. My issue is that I am creating a Top-Down game and the player will never see more about 50 feet in any direction. Currently, all meshes and foliage are forced into their low poly versions before they are off the screen. In a Top-Down view, this looks really bad.

My questions are:

  1. Can I disable LOD entirely?
  2. Are things rendered off screen at some X distance, before they are shown on the players screen? If that is the case, can I lower that draw distance?
  3. What map sizes would you recommend for a Top-Down view? My map will be one large seamless world, with transitions to dungeons, etc. Unless this is highly frowned upon in UE, but I am reminded of Fortnite/PUBG, so I know it is doable.
    3.1) How would a tiled map compare, for performance, etc?

Is your camera completely fixed or can it zoom in/out? If it’s completely fixed, then you can probably disable LODs.

How fast can your character move/camera pan? If it’s not 100s of meters per second, you can probably do some tile streaming with some math to compare where the camera is vs the bounds of the current tile and then start loading the next tile nearest.

You’ll want to avoid any lengthy streaming popins, so making sure to preload the next nearest tile is going to be your best bet. I’d try to take your camera frustum and triple it for tile size. So if your camera view is roughly 30m wide, make your tiles 100m(rounded up for ease of integration) wide. That would give you plenty of time to preload the next tiles.

All of this can easily be done with sub levels. Your persistent level would be what would hold the basic world collision so that things like AI don’t break.

I copied all of that down and stored it in my notes. I have no idea what you said so I will need to do some more research. Thank you.