3 Advanced Simplygon Tools to Improve Your Automated Character LOD Pipeline
An automated LOD pipeline can save time and ensure consistency. Here’s how to use Simplygon to take yours to the next level.
In modern game development, Level of Detail (LOD) optimization is essential for maintaining performance without sacrificing visual fidelity. If your game features complex characters, an automated LOD pipeline can save time and ensure consistency. In this post, we’ll explore three advanced techniques using Simplygon to take your character LOD pipeline to the next level.
Keep topology with quad reduction
For LODs close to the camera, preserving the original topology is critical for animations and skinning. With Simplygon’s Quad Reducer the topology is kept close to the original asset. This is also a good tool to use to generate variations targeting lower end platforms, so called LOD0 optimization.
Reduce modular characters with modular seams
Modular characters allow for customization, but they introduce challenges when generating LODs. It is often the case that the parts does not line up after optimization. Modular Seams in Simplygon ensure that LODs for modular parts align perfectly after optimization, preventing visible gaps or mismatched normals.
Traditionally modular characters have been handled by hiding the seam in different ways, or not optimizing the seam at all. Each of these workarounds have their own issues. With modular seams, you can create LOD levels for character parts without having to worry about gaps.
Simplified shaders for distant LODs with vertex color casters
For distant LODs, texture sampling can be expensive. Vertex Color Casting allows you to bake color information directly into vertices, reducing texture memory usage while maintaining visual fidelity at a distance. Using vertex color casting enables all distant LODs to share the same material.
Final thoughts
These three tools can enhance your automated character LOD pipeline, ensuring high-quality assets while optimizing performance. To get the full usage out of them you want to have a fully automated pipeline up and running. Our advice is to solve one step at a time, always benchmarking performance gains. Two good blogs on this topic are: