First -- thank you to all the encouraging people out there & in our discord who are enjoying the game where it's at and encouraging us to bring it to its potential. It's been a little bit since we've pushed an update to the game and that's because we made some bold choices to get the game where it needs to be. Taking in people's experiences and looking at what we need to do to bring it to completion, we landed on a few aspects that need extra focus. Some of which are optimization, stability and performance.
So, we tore out some insides which makes almost everything not work until we get it back together. Horrifying? Yes, but very worth it. Now we're in the process of upgrading a few key aspects like the terrain, how we handle detail levels of objects and surfaces and technical aspects of the vehicles' structure. These should make a high impact. I'll go into a little detail on some of these below for those interested.
Our next goal on the horizon is to get the upgraded version into a stable update for the game - the next post I do, as soon as it's ready, will detail what we've upgraded exactly. It will be posted along with a game update that includes most of the below changes & optimizations, once they're done.
While we aren't expecting insane miracles in terms of performance increases, initial results are very promising. We're taking a deep look at what optimizations we can make without sacrificing key gameplay and executing what is reasonable, prioritizing non-loss optimization. From this, we determined a few areas that could get the largest optimizations: Terrain, object rendering detail and vehicle structuring.
{ Terrain Rendering }
We redid the rendering system for the terrain, giving us more control as well as more speed in rendering. This cut a major base cost of performance and while we're having to rebuild certain aspects of the terrain's rendering system, the performance benefits were significant. We unified the terrain's visuals with object and plant visuals, which has a performance benefit to rendering and strain on the graphics card but it also let us put things like wind, wetness, dust, snow all into the same system of rendering. Which opened up much easier paths to implementing things like rain, snow, dust storms, etc.
{ Soft Terrain }
We've improved how vehicles interact with non-solid parts of the terrain, like mud, sand or snow. We have deeper control of the physics of the soft tires interaction with "soft" parts of the terrain - this will directly improve the experiences of getting stuck (or not) in mud or sliding in deep sand. It's possible further performance benefits can be gained from the dynamic terrain system, we'll see.
{ Object Rendering and Collision }
We have gone through about 75% of the objects in the game, optimizing them individually to meet a new unified standard we designed to allow for a spread of graphics quality options as well as optimized rendering across all options levels. Improvements to them include unifying their texture rendering method, swapping low quality/needlessly heavy models for more reasonable ones, making sure every object is uniformly setup with various levels of detail to help with lower end graphics settings as well as distant objects not needlessly rendering detail and lastly, making a system to determine the needed collision detail level of objects in the areas based on their size to keep the collision on objects like rocks precise everywhere that it is needed and looser where it's not.
{ Vehicles }
Lastly, we found ways to structure the vehicles that will allow them to save possibly over 50% of their base performance cost per active vehicle. This is huge for optimization but also in terms of multiple vehicle support, it makes it more likely we'll be able to have multiple vehicles be a part of gameplay instead of restricting to one active vehicle at a time. We'll see how it comes together but either way, I'm happy with the way vehicle optimizations are shaping up.
Thanks for reading! We're excited about the optimizations and solidification of systems we're using for the game. It feels like a drastic refinement - we didn't exactly expect to need to do it but.. once we identified it as required and started in on it, it's proving to be the correct choice.
