No announcement yet.

Double-Precision Positionning support WIP : Continuing Effort thread

  • Filter
  • Time
  • Show
Clear All
new posts

    Currently using Origin Rebasing you can expand your worlds to the box of +-21 000 km in each direction.
    I've developped a library of unreal engine vectors, transforms, rotations mostly converting FVector, FVector2D, FVector 3, FTransform, FQuat etc... to use and operate using double precision.

    You first need to whange the Unreal Engine 4 Variable WORLD_MAX to go beyond the base +-20km of playable area.

    LinkedIn | Link custom Shaders | Atmosphere Modelisation & Procedural Planets | Distance Matching Locomotion | Nvidia GameWorks builds - 4.19.2 : VXGI2.0, Blast, HairWorks, Flow - Plugins: VictoryBP


      I came across this trying to figure out at what distance relevant accuracy is lost. I was wondering, and realize do to other components such as Physics engines and graphics cards etc. still use 32bit float that 64 bit float most likely will not be implemented? I am curious about making an RTS that runs as a sim for calculating actual projectile physics etc. and map sizes ranging from 5km to 80km+. I see where people often sat 32bit float gives enough precision, but in my research I am finding otherwise, as devs mention losing accuracy at even 1km distance from origin. I want accurate simulation for a multiplayer game (World Comp, level streaming etc. does not support multiplayer) at ranges beyond 80km. After all, real world Artillery can fire rounds 23km! We are so far in technology, what is it going to take for all systems to switch to using 64bit float?


        Faced with the problem of limiting the size of the world. Can you tell me which engine sources need to be modified to switch from float to double ? need network support as well

        it is clear that I need to change vector.h in the first place,

        what else ?
        Last edited by aGeronimo; 10-18-2020, 03:18 PM.