raytracing uses bounding boxes and geometry clusters aka bvh, while sdf is raymarching in basicly screenspace. combining those is i’d guess rather expensive or maybe not compatible. cause screenspace and bvh don’t share data. or you would raytrace in world space and if you hit a bounding box of a sdf volume you would start to raymarch. not sure if this is applicable or if shaders can switch that style of code. i’m not that deep in the engine. i can visualize it, but not code it very well, beyond hobby level. : )