>As for code itself, Carmack implemented the pre-warp as a shader that projected the rectilinear image onto a surface (geometry). This approach was abandoned ye

>As for code itself, Carmack implemented the pre-warp as a shader that projected the rectilinear image onto a surface (geometry). This approach was abandoned years ago and is now implemented with a pixel shader. Not that it matters for your point, but it was the other way around. They are using a mesh based pre-distortion now. >Although the exact distortion parameters depend on the lens characteristics and eye position relative to the lens, the Oculus SDK takes care of all necessary calculations when generating the distortion mesh. https://developer3.oculus.com/documentation/pcsdk/latest/concepts/dg-render/ https://developer3.oculus.com/documentation/pcsdk/latest/con...