Storing color data in the vertices of our meshes allows the PBR shader to run an efficient, client-side linear multiplication of that color over the Albedo/Base Color texture map. This is incredibly useful for breaking up macro repetition on trim sheets and tiling textures across large builds in a far more precise way than the full material slot tinting we have now. It allows creators to reuse materials more heavily without resorting to unique bakes for every single wall or structural chunk, mitigating excessive decal use and secondary macro-variation layers. This workflow is ideal for adding fine artistic control, such as deeper ambient occlusion shadows in crevices, highlights on edges, or entirely new stylized aesthetics (like low-overhead cartoon texturing) that rely primarily on vertex colors for variance & styles. Key Considerations Texture Memory (VRAM) Savings: Texture memory consumption is a primary source of client-side viewer lag and texture thrashing. By allowing vertex colors to drive object-based ambient occlusion, gradients, or unique room/object shading variations, creators can aggressively reuse a very small set of tileable materials or trim sheets while maintaining visual variety. Without this, creators frequently tend to generate massive, unique texture sets for every single wall and asset variant simply to display localized baked ambient occlusion & variations. Reduction in Alpha Pipeline Bottlenecks & Mesh Bloat: Currently, creating macro shadows or grimy variations requires duplicating face geometry shells to use as decals or AO layers. This forces the engine to run overlapping draw calls and increases geometric complexity. Furthermore, if creators use alpha-blended textures on those shells to soften shadow edges, it triggers expensive depth-sorting calculations which can tank viewer frame rates. Moving this occlusion data directly into the vertices removes the extra mesh layers & textures entirely and completely eliminates alpha-sorting overhead. The glTF Pipeline "Easy Win": Because Second Life's rendering engine is now anchored to the glTF 2.0 specification, the underlying framework for this already exists natively. The mesh importer simply needs to retain the COLOR_0 attribute data during processing instead of stripping it out, and the PBR shader needs an efficient vertex color multiplier step applied to the base color factor. Perhaps if you choose to move forward with this you'll also consider some Object Properties panel settings, likely an on/off checkbox for vertex color effect and a slider to tweak its power would be more than enough. Maybe even opening this up to scripting could prove interesting for creating all sorts of different techniques and effects that leverage vertex colors. For Vertex color you get the full RGB spectrum to multiply over the albedo, but that still leaves the alpha channel you could potentially use for other things, like adding emissive, or even going as far as allowing a two material texture blend shader. Using the alpha channel to control emissive would be super light and efficient, but material blends can get heavy if not optimized properly, still it is possible lol, so I figured I should add it in context. Community Discussion Reference: I made a community thread to gauge interest and feedback, which can be found here: https://community.secondlife.com/forums/topic/534987-any-thoughts-on-adding-vertex-color-multiply-to-the-new-pbr-shader/#comment-3097036 Just to be clear this is no demand nor expectation! Its just an inspiration I had when considering you all are working on performance boosts and adding new features for creative potentials, I feel like this could be an easy win in both regards.
