[[[email protected]]] - > Image distortions in the slow scan direction are clearly reduced when using Hilbert scanning, as vertical streaks in the calculated FT were not apparent. However, drift of the sample/stage is still present and results in abrupt discontinuities in the image where regions of the image that were taken with a longer time between them meet. This effect increases with increased dwell time, indicating that indeed this issue is related to slow sample drift rather than settling time issues with the scan system. Figure 6 shows these abrupt changes in images acquired at different dwell times. Indeed, rather than a line modulation as in raster and snake scanning, the modulation due to sample drift is now smeared over the image with approximately equal weight in all directions, leading to a better behaviour when calculating the diffractogram. - > The total standard deviation caused by drift in terms of strain mapping is however nearly independent of the scan method and one could see this is as an ‘error-budget’ that is more equally distributed over x and y direction for Hilbert scanning as opposed to the two other scan methods. > There are however two important benefits in doing a single Hilbert scan as compared to two orthogonal raster or snake scans. > (i) Single scan is faster and requires a lower dose, which might be essential for beam sensitive samples. > (ii) A single scan containing correct information about x and y strain allows to correctly calculate the relation between both strain directions which is important for e.g. shear strain. This is more difficult for raster scanning as it requires combining low noise x and y strain maps from two different images. - > In terms of dose, both snake and Hilbert scanning are preferred as no flyback time is required (unless a fast beam blanker is available in which case the raster scanning will provide the same dose). The distribution of the dose with the Hilbert method is clearly different than with the raster method. The Hilbert method registers small neighbouring areas of the sample in a shorter time than line by line scanning, this could have an influence on beam damage as the dose is distributed faster in smaller areas, further experiments need to be performed on beam sensitive samples to clarify these effects. In the field of selective laser sintering, it was shown that the distribution of temperature in the treated object was more uniform when scanning with a Hilbert method compared to a raster method