- Jun 2, 2012
- 6,470
- 32
- 91
Every game that features large maps has set draw distances on everything from which models to use, which textures to use, which shadow resolution to use etc... because computers can't render the whole map in full detail. The problem is it breaks immersion when you see it happening before your eyes which has become all too apparent in the post-programmable shader era in which the amount of up-close detail has increased tremendously but distance detail, while increasing, hasn't increased at the same pace and is still using the same technique of flipping between 2 or 3 different LOD models as the character/camera gets closer to the object.
Is there any solution to this rendering technique, or is graphical pop-in here to stay? Could tessellation be used instead to more discretely ramp up the detail on objects in a less obvious/intrusive manner, basically replacing the use of different LOD models? Or is tessellation way too expensive to be used in that manner?
Is there any solution to this rendering technique, or is graphical pop-in here to stay? Could tessellation be used instead to more discretely ramp up the detail on objects in a less obvious/intrusive manner, basically replacing the use of different LOD models? Or is tessellation way too expensive to be used in that manner?