The brightness and colors of stars are fundamental data in studying the structure, size and age of the Universe, and in testing theories of stellar structure and evolution. But such measurements can be degraded by large systematic errors; current state-of-the- art techniques are less accurate than often sopposed, leaving basic astrophysical conclusions open to question. A fundamental barrier to further progress is that existing photometric systems violate a basic requirement of information theory, the sampling theorem. This award will support the development of a correctly sampled photometric system, and demostrate that it produces results of greatly improved accuracy, not only reaching the accuracy that was expected of undersampled systems in the past, and needed for general astrophysical work, but even surpassing this and making new investigations possible. Because correctly-sampled photometry can be reduced to a standard system without loss of accuracy, regardless of detector type, such data taken with all detectors will transform to a common system without systematic errors. This technique may let ground-based observers match the Space Telescope filter systems to arbitrary accuracy, which cannot be done with conventional methods.