Release notes

Version Mar 30, 2022

• Correction of wrongly reported spot coordinates in file SPOT.nXDS for spots comprising a singel pixel (see MINIMUM_NUMBER_OF_PIXELS_IN_A_SPOT=)
• Exchange of a few instructions and type declarations to make the source code of the nXDS package strictly obeying the Fortran 90 language definition.

Version May 10, 2021

• Replacement of the spot finding routine in the COLSPOT step by a more robust procedure.
• The new input parameter   MAXIMUM_NUMBER_OF_SPOTS= defines an upper limit for the most strongest spots accepted by COLSPOT in any image.
• New input parameter GAIN= allows to fix the conversion factor between image pixel contents and equivalent counts (for a pixel detector you could set GAIN=1.0).

Version Feb 14, 2019

• Substantial revision of CORRECT/nXSCALE involving new data structures.
• Addition of a new processing step, FILTER, for rapid recognition of useless snapshots.
• Expansion of compressed images by nXDS has been restricted. In case an external library routine is specified by the input parameter LIB=, expansion of compressed data is now up to the user supplied reading routine.
• Switch back to the original expression for the Lorentz correction factor for 'still' snapshots as given in reference Kabsch, 2014. The change to the inverse form of the correction factor, introduced with the BUILT=20180508 release, was caused by a bias in the nXDS calculations that could be compensated by inverting the Lorentz correction in some cases. The bias is visible predominantly at low resolution when reflections are sampled by stills mostly at their tails (Gleb Bourencov, gleb@embl-hamburg.de).
• Increased maximum of the allowed number of connected files to avoid potential restrictions in the use of parallel threads.
• Improved method for determination and refinement of mosaicity. In the INTEGRATE step all reflections are saved now which are within 2 estimated standard deviations of the rocking curve off from the Ewald sphere.
• The input parameter MINIMUM_EWALD_OFFSET_CORRECTION= is now only used by the CORRECT step of nXDS and by nXSCALE and has a new default value of 0.4.
• The parameters  FRIEDEL'S_LAW=, and  MERGE= are no longer used and ignored.

• Improved method for estimating intensity variances based on observed intensity values of symmetry-related reflections.
• The main reflection output file nXDS_ASCII.HKL has been shortened. Equivalent reflections from all images are randomly assigned to either one of two representatives in each parity class of Friedel pairs and the weighted mean intensity of reflections contributing to each representative is returned.
• A new optional input parameter LONG_OUTPUT_FILE= is provided that contains the individual reflections from each snapshot.
• Information about anomalous effects is no longer printed by the CORRECT/nXSCALE step. For unknown reasons false positive 'significant' anomalous signals have been reported in some cases that could lead to confusions. The output file nXDS_ASCII.HKL always keeps Friedel pairs separate - leaving detailed analysis to other programs.

• Change in the default setting of the input parameters MERGE_TREE= and POSTREFINE=.

• Correction of a rounding problem in the IDXREF step that previously caused unnecessary program termination with the message !!! ERROR !!! ASSERTION #1 IN "MERGTREE" (reported by Aaron Finke).

Version Jun 25, 2018

• Extensions for dynamic linking of an external image reading routine. This renders nXDS more independent of the technical details of future detector image formats. All image formats accepted by XDS are now readable by nXDS as well. As described in section Images this required the introduction of the new input parameters
  NAME_TEMPLATE_OF_DATA_FRAMES=,  DATA_RANGE=,  EXCLUDE_DATA_RANGE=,  LIB=.

• Bug correction in case of almost empty snapshot images that had caused nXDS to collapse in the indexing step. Bug reported by Jasper van Thor.

Version Jan 30, 2018

• Format change of file nXDS_ASCII.HKL. In case of unmerged final output data each reflection record has now attached 2 numbers identifying the input file (type INTEGRATE.HKL) and the image number within the input file where the reflection was recorded.
• nXDS now handles images obeying the definition of TIFF Revision 6.0

• Bug correction in allocating snapshot file names.

• The new input parameter MERGE_TREE=, effective in the IDXREF step for unknown crystals, allows the user to select translated reciprocal lattices for merge with the dominant one.
• The input parameter SEPMAX= used previously has been removed.
• Improvements in speed and accuracy for locating difference vector clusters in the IDXREF step.
• The previously used expression of the Lorentz factor for 'still' snapshots was shown wrong by Gleb Bourencov (embl-hamburg). The correct form now used is the reciprocal of expression (42) given in reference Kabsch, 2014.

Version Oct 14, 2017

• First release of nXDS. The package is based on the rotation data processing program XDS, version June 1, 2017.

• Format change of file INTEGRATE.HKL Each reflection record has now attached the running number of its image.