Final Activity Report Summary - APCOCOS (Appearance prediction of colour coatings based on under-surface scattering and optical measurements of bi-directional reflection distribution function) The field of predictive rendering concerns itself with those methods of image synthesis which yield results that do not only look real, but are also radiometrically correct renditions of nature, i.e. which are accurate predictions of what a real scene would look like under given lighting conditions. A real coating consists of pigments, effect pigments, clear lacquer and glaze. A novel and unique combination of real parameters that are commonly measured in the industry and a theoretical reflectance model consisting of measurable parameters was proposed. The authors designed perception parameters and put them into well-known surface reflection functions such as He and Torrance. The original contributions are the study of the sub-surface scattering of real paint and the prediction of its appearance in rendered images by the proposed model of light reflection beneath the paint surface. The other contributions are the high dynamic range (HDR) sky model above the city Sred that has been developed and used as natural illuminator of our paint surfaces. Because many paint producers want to keep their chemical compositions secret we have developed the concept of Black box paint model that keeps all the measurable parameters in its database related the reference ID of paint. By this way all parameters needed for visual prediction are available and the composition is not opened to public. A new objective established during the course of the work was the observation of sparkling effect in some paints that has been captured by HDR static photos and visualized on common monitors and HDR monitors with comparison of the visual quality. Surprisingly, the best visual appearance of sparkles has been observed on common monitors.