INTERFERENCE COLORS IN MINERALS AND DIGITAL COMPENSATION

Birefringence links interference colors to refractive indices and retardation. Compensation measures retardation. Color is a complex attribute to render faithfully. The math needed to predict the birefringence of a mineral plate was developed in the last century but extracting useful numbers by hand is tedious. Computers brought both problems under control. Routines that return useful numbers often can be implemented on a spreadsheet. Computers have transformed the handling and management of color. Digital renditions of the Michel-Levy chart can be created, viewed, and printed. Retardation can be derived from interference colors by: (1) Visual comparison of an interference color with colors on a Michel-Levy color strip. (2) Computer-aided color matching of a photomicrograph with a list of digital values for RGB components of light. (3) Inserting a compensator in the accessory slot of the microscope. Precision gets better from method (1) to method (3). Method (1) is the most commonly used. Most people can detect a difference in interference color of 100 nanometers (nm) or less on a Michel-Levy chart. Pick a color on a Michel-Levy chart and most can see a different color within 50 nm in either direction. Given a standard thin section 0.03 mm thick, 50 nm translates into a birefringence of approximately 0.0016. One cannot detect birefringence differences quite this small among some of the first order grays or among the colors of fifth and sixth orders and above. But in the range of interference colors displayed by many rock-forming minerals, those with interference colors corresponding to retardations between 250 nm and 1650 nm, one can. A colored photomicrograph that faithfully represents the colors displayed under the microscope can be used to infer a retardation from the interference color with a precision of a few nm or less (method 2). A precision of 1 nm translates into a precision for the birefringence of 0.00003 in a mineral 0.03 mm thick. Like method (1), method (2) cannot detect birefringence differences this small among some of the first order grays or among the higher order colors. The method fails if the interference colors are anomalous. Precise birefringence values constrain estimates of olivine compositions and pyroxene orientations in thin section.