Abstract

Recent studies¹ have indicated that an array of microlenslets with a near-unity fill factor can be used to direct light more efficiently onto an underfilled array of solid-state photosensors. Radiometric gain may be defined as the total number of rays, directed onto a photosensor area by a lenslet, divided by the total number of rays collected by a bare photosensor, assuming the same ray density. We present the results of ray-based radiometric calculations which model the collection of light by a typical objective lens onto a composite imaging array. This array consists of three optically important layers: the microlenslet layer, a transparent spacer layer, and the detector plane. The radiometric gain is presented as a function of spacer-layer thickness, f/No. of the taking lens, image spot size, and microlens cap height. Radiometric gain values exceeding three are predicted for a typical video imaging array when the lenslet is modeled as a spherical cap truncated to match the pixel area. A lower gain is predicted when an anamorphic-pillow model is assumed. Experimental verification of the spherical gap model is demonstrated for lenslets formed by baking rectangular pads of photoresist. A departure from the model's prediction at large cap heights is explained by a change in the lenslet’s shape.

© 1986 Optical Society of America