Hi
I’m writing a script that will do the following:
Using a dummy scene that consists of a matrix of tunnels (mining related) the script will be a movebale (via controller) GO that can fly around the scene. When an assigned button is pressed this will effectively sow or seed GO’s in the scene i.e. small rocks & debris objects. Basically a seeding machine that drops objects as I move around like spraying crops. There will be additional parameters to randomise sizes etc. to ensure a variant scatter.
As the GO’s are seeded they will intially have physics rigidbody attached such that they fall the ground (terrain) and settle. The code will then wait (IEnumerator) a second to allow physics to setlle the then the rigidbody and collider will be removed from the GO levaing it static.
Once the objects have been sewn … the next part of the script will use a free asset called “Mesh Combiner”
Each of the sewn objects (single material) are iterrated through and are then combined to make a single mesh prefab of many smaller objects. These will then be used in the LIVE scene.
MY KEY QUESTION:
Will combining many objects over a wide area have any significant impact on performance? Or should I group by localisation i.e every 30 m or so? Imagine a tunnel 100m long 5m wide in which I want to scatter debris along and combine such for perforance on draw calls. All objects (in this case) will use the same texture Atlas. My guess is about 10 objects per meter (very low poly)
I dont know if billboarding or using tree planter on terrain might be a better suited idea? These object will not have to have any physics or colliders its just asthetics to the environment.
Many thanks
Archy