So Im looking to create a game comparable to Stonehearth, though without blocky asthetics - not necessarily for sale, more to expand my horizons and knowhow. After considering Unitys Terrain tool, I found it unfit for my grand designs and am tentatively coming to the conclusion of making my own Terrain module/tool. One question is keeping me up at night though:
How do I handle LOD?
Im using .tif grayscale heightmaps for now and Ive gone through a few ideas but none of them really “feel” good …
Im reluctant to simply rely on Tessellation to get the job done, seeing as how the entire map in full resolution would need to be loaded, bringing obvious drawbacks. I thought of generating separate LOD resolutions to complement the Tessellation (range-dependant), mitigating many drawbacks. This would create an overlap problem with the separate LOD resolution maps at LOD border regions though, not to mention the potential problems with verts in the transition zones of the separate LOD resolutions…
There’s plenty of high-class, high-value terrain assets available in the store. Be sure to evaluate those as well because you might end up re-inventing the terrain-wheel where you might have saved yourself upwards of a thousand hours of development time in exchange of probably less than $200.
Oh definitely - something like Digger Pro looks like it would work nicely for what I have in mind, not even 100$ - Im not looking to make a sellable game though, its mostly for my own benefit/experience and buying a tool would somewhat defeat the purpose. Besides, it might look like it does what I would want - but I have no way of knowing without trying it out and potentially finding it useless for my purposes. From what Ive read about Unity Terrain, I dont have much faith in it being able to do what I would like of it, nor do I think I can influence it meaningfully
To clarify the scope a bit, I’m looking to find a way to merge traditional poly/heightmap techniques with real-time in-game editing (which Digger appears to be able to do, though again, I have no way of knowing the malleability of the tool itself). With that, Im trying to further my experience with manipulating meshes in-game and generally experiment or implement tried and true methods. Specifically that last aspect is the main reason for posting.
You google “adaptive terrain rendering” and implement one of the algorithms. People kept tinkering with this since 2000s, there should be plenty of reading material right now.