The optical properties of glazings and coated glazing products at oblique angles of incidence are discussed. The parallel shift of both transmitted and reflected beams due to refraction and multiple reflections within the sample poses problems for most standard instruments. A procedure for the correction of incorrectly recorded signals due to this parallel shift of the back surface reflected component is presented. An absolute instrument using an integrating sphere as the detector is described. This instrument is insensitive to a parallel shift of the incoming beam and thus detects all transmitted and reflected beams from transparent glazings at oblique angles of incidence correctly. Model calculations for the prediction of the angular variation including the Fresnel formalism as well as empirical algorithms are discussed. Comparing the results of exact Fresnel calculations with the results of calculations using such an empirical algorithm shows that a simple polynomial with only two terms can predict the angular variation function of coated glazings accurately, provided the polynomial is chosen in accordance with the coating category being evaluated. Because of the low precision of experimental results and the complex nature of exact Fresnel calculations, I suggest that such an empirical algorithm should be used as a standard procedure to predict the angular variation of reflectance and transmittance of glazing products. This algorithm would lead to simplified procedures and more unanimous results than those that are achieved today. (C) 1997 Elsevier Science B.V.
