(a) Samples from run #83 were measured by three additional laboratories. follows. The results are as Note: All samples are deposited on super polished F.S. substrates with RMS roughness <0.75A and no visible defects in central aperture. Measurements of total loss and scatter are difficult at few ppm levels. The dependence of the measurement depends on the condition of the components and on the spot size. These can cause great discrepancies in the measurements, making it difficult to make conclusions on the reliability of each measurement. However, the measurements presented in this report were found valuable in development of very low-loss mirror coatings. The measurements are done at the 632.8nm wavelength and are limited to parts with less than 5 X 10-3 total loss. total loss. In the case of the loss measurements, the greatest virtue of the two instruments is the simplicity of the actual measurement and the very low cost of assembling the two measurement setups. 9. References [1] Anderson, D.Z.; Frisch, J.C.; Masser, C.S. Mirror reflectometer based on optical cavity decay time. Applied Optics. Vol. 13, p. 1238; 1984 [2] Yariv, A. Introduction to Optical Electronics. Holt Reinhart Winston. p.p. 98-140; 1971 The Art of Optical Scatter Measurement Thomas A. Leonard Ball Systems Engineering Division Fairborn, OH 45324 (513)429-5005 Manuscript Received A BRDF round robin was recently conducted in the United States with 18 participants from government, university and industry scatter laboratories. The spreads in reported BRDF levels from the set of four master samples were quite large. They clearly show the need for more rigorous measurement procedures within the scatter community. Suggestions are made for optimizing measurements and minimizing errors, based on information gathered during the round robin. A "BRDF standard test method" being considered in ASTM subcommittee E12.09 is discussed. KEY WORDS: BRDF; Round Robin; Scatter 1. Introduction Optical scatter measurements are subject to all the subtle inaccuracies and complications that plague any radiometric measurement. A recent nationwide BRDF round robin confirmed that the scatter community is not immune to these problems. Large differences in reported BRDF levels from a set of four master samples clearly show the need for more rigorous measurement procedures and the utility of physical reference standards. The author was present for all but one of the 18 measurement sets in the round robin. A good deal of information was accumulated on instrumentation, procedures and capabilities which should benefit the entire scatter community. Standardized measurement procedures and data formats would be very helpful for data intercomparisons, and to that end we are pursuing a standard test procedure through ASTM. Possible continuation of the round robin measurements and promulgation of a standard test procedure will hopefully bring the scatter community closer to the accuracy levels needed for critical design of space-based optical surveillance systems. The master sample set for the round robin consisted of four, two-inch diameter samples. In order of decreasing BRDF: W B M Aluminum disk coated with Desothane Aliphatic untinted white Polyurethane enamel for a diffuse white surface with a flat BRDF of about 0.27/sr at 633 nm. Aluminum disk coated with gunship black Polyurethane enamel for a diffuse black surface with a BRDF of about 0.01/sr at 633 mm. Industrial grade, bare Molybdenum mirror with a surface roughness of about 45 A rms (40-1000/mm). A - Aluminized Sio coated glass mirror with a surface roughness of about 7 A rms (40-1000/mm). In-plane scatter at 0.6328 μm and 10° angle-of-incidence was measured from these samples at 18 optical scatter facilities in the United States. They included government laboratories, universities and industry. A detailed report on the round robin was presented at SPIE Conference 967, "Stray Light and Contamination in Optical Systems", San Diego, August, 1988. Figures 1 and 2 show the BRDF results for the W and A samples. The reported curves from this diffuse and specular sample are sufficient to demonstrate the problem areas in BRDF measurements. In an effort to quantitize the variations, a statistical analysis was made of the 18 measurement sets for each sample over the range of |