- RegisterQD now requires an extra step because of enhancements that support either CPU and GPU processing. For CPU processing (formerly the only option), now you must manually load the RegisterMismatch package:
using RegisterMismatch, RegisterQD. For GPU processing, you should instead load the RegisterMismatchCuda package:using RegisterMismatchCuda, RegisterQD. Note that loading both mismatch packages in the same session will cause method conflicts. Both mismatch packages are registered in the publicly-available HolyLabRegistry, and users are advised to add that registry.
- For users of
SD("space directions"), an input that allows you to specify the spatial sampling of your input array, the meaning of the returned transformation has changed. The return value is now in "physical space" rather than in "array index space." Specifically, formerly the returned transformation included the consequences of any change of scale needed to apply the transformation to the supplied arrays: for a rotation matrixR, the return wasSD\(R*SD). Now justRgets returned. This means that a rigid transformation will now "look" rigid. This change fixes some formerly-significant problems for optimization when a known starting guess was available. Use the newarrayscalefunction just before warping to prepare a transformation for a particular array with nonisotropic sampling.