Abstract: A system and method are used to increase alignment accuracy of feature patterns through detection of alignment patterns on both a surface layer and at least one below surface layers of an object. Visible light is used to detect alignment patterns on the surface layer and infrared light is used to detect patterns one layers below the surface. For example, reflected visible light and transmitted infrared light are co-focused onto detector after impinging on respective alignment patterns. The co-focused light is then used to determine proper alignment of the object for subsequent pattern features. This substantially increases accuracy of alignment of pattern features between layers, as compared to conventional systems.