Abstract
Natural materials cover a wide range of surface scattering properties, which can be described by the bidirectional reflectance distribution function (BRDF). The lobes of the BRDF may be described by their primary scattering direction (resulting in salient features of the appearance modes at different locations) and generic BRDFs typically consist of several of such lobes. Thus, similarly to describing BRDFs as linear superpositions of scattering lobes, we can describe the appearance of objects consisting of any materials as linear superpositions of objects of different canonical materials. We approach this problem via optically mixing of objects that represent canonical reflectance modes, allowing systematical and gradual variations of materials to test perceptual qualities of generic materials. We performed two experiments with a shiny, matte and velvety green bird, representing the forward, diffuse and surface scattering modes. The mixing was done with combinations of two materials in a real setup, using a semi-transparent mirror, and with combinations of three materials on a screen, using superpositions of photographs. In both experiments we tested how observers rated glossiness, softness, warmth and heaviness for systematically varied weights of the materials combinations. Finally we conducted a survey about how observers experienced the optical mixtures. We found systematical, gradual variations of the ratings as a function of the weightings, for the real stimuli consisting of dual mixes and for the images consisting of triple mixes. The data patterns for the real stimuli and for the images were consistent with each other for glossiness, softness and warmth ratings, but showed a global scaling and slight local modulation for heaviness. Observers thought we were varying material, lighting and color and they found the birds to look real. These results show that we can use optical mixtures as a basis for a material probe to investigate perception of generic, real materials.
Meeting abstract presented at VSS 2014