Figure 5
Plots showing how the dark-field parameters vary with sample-to-detector distance d for uncorrected and corrected data, legend shown at bottom of the figure. Semi-major blurring width \(\sigma _M\) (A) and scattering angle \(\Theta _M\) (E) for carbon fibres. Semi-major blurring width \(\sigma _M\) (B), semi-minor blurring width \(\sigma _m\) (D), and root-mean-square scattering angle \(\Theta _{RMS}\) (F) for microspheres. The data is the mean value of the parameter extracted from the regions in (C) highlighted by the green boxes. The retrieved blurring widths should increase linearly with sample-to-detector distance, which is not the case for the blurring widths retrieved from the uncorrected data in (A), (B), and (D), while the blurring widths retrieved from the corrected data appear more linear. The scattering angle \(\Theta\) should remain constant with sample-to-detector distance. Note that the error bars on these measurements are smaller than the data points.
