I am making a first person game in a haunted house.
Is it better in terms of the graphics performance to load meshes just before the player (camera) can "see" them, for instance just before a player enters a room where these objects are located, and then use something like FreeEntity to dump them again once the player has left and can no longer see them, or do the number of meshes/textures etc. only have an effect on performance for the duration they actually appear on the screen anyway. In other words, do meshes/textures that are loaded but currently hidden from view slow down the game? (Assuming they are inanimate like furniture etc.)
In addition to inanimate objects like furniture, I have a doll that drifts through the air randomly in one of the rooms, will the code responsible for making the doll float around/change direction slow down the game even when the doll can't be seen? Should I make it move only if it can be seen (using EntityVisible perhaps)?
Is it better in terms of the graphics performance to load meshes just before the player (camera) can "see" them, for instance just before a player enters a room where these objects are located, and then use something like FreeEntity to dump them again once the player has left and can no longer see them, or do the number of meshes/textures etc. only have an effect on performance for the duration they actually appear on the screen anyway. In other words, do meshes/textures that are loaded but currently hidden from view slow down the game? (Assuming they are inanimate like furniture etc.)
In addition to inanimate objects like furniture, I have a doll that drifts through the air randomly in one of the rooms, will the code responsible for making the doll float around/change direction slow down the game even when the doll can't be seen? Should I make it move only if it can be seen (using EntityVisible perhaps)?