The `Seed-Skewness' Method for Integration of Peaks on Imaging Plates II. Analysis of Bias Due to Finite Size of the Peak Mask and Treatment of α₁-α₂ Splitting

The conditions for the skewness minimum in the seed-skewness algorithm for integration of diffraction maxima on imaging plates are examined. Theoretical estimates of the bias in the integrated intensities due to finite size of the peak mask are derived, and compared with experimental values based on measurements over a single oscillation range with different experimental parameters. The bias is present not only in the seed-skewness method, but also in the standard-box integration results, which suggests that it is at least in part a result of detector characteristics. Appropriate corrections are needed if the data are to be used to the fullest extent in high-accuracy applications. An algorithm for the intensity correction in the case of splitting of the reflections is described, and tested by analysis of a data set collected at a rotating-anode source. The analysis shows that the most reliable α₁−α₂ correction is obtained by multiplication of the α₁ intensity by an a priori determined factor, and that, on merging of long and short exposures, the weak peaks from the short exposures should be discarded in both rotating-anode and synchrotron data sets.