Choosing wavelengths and assessing blunder risk for the method of exact fractions

We revisit the problem of choosing combinations of sources for multi-wavelength interferometry, using the risk of assigning the wrong integer interference order to a measured phase as a quantitative criterion. We model this risk, explicitly accounting for the experimenter's ability to reject implausible measurement results. We present a general reliability limit for the method of exact fractions, which can be used as a benchmark when evaluating wavelength combinations, as well as performance estimates for several promising combinations.