The SMC application can be coupled with Fit2D to offer a range of data reduction and manipulation features for large file series. The program supports batch-processing via the compilation of simple point-and-click file processing lists, allowing non-sequential file names and multiple data paths. In this way, up to 60'000 files can be batch-processed in each user session. The software features multiple background subtraction modes, including sequential subtraction for rapidly identifying differences and same-pattern subtraction for the removal of isotropic contributions from anisotropic features. For single file subtraction, different background files can be automatically associated with individual file groups. This permits the use of up to 30 unique background files in each session. This can be used to process multiple data-sets together, or provides a way of solving the problem that numerous background files may be required at different stages within single large scans.

As with Fit2D, the SMC application can be used to batch-produce a range of different output types, from intensity-corrected and cropped patterns to 2D matrices. However, the most frequently employed data reduction process is the integration of area detector data into radial or azimuthal scattering intensity profiles. Consequently, this is particularly well-supported by the SMC application.

In the SMC utility, integration limits can be set either graphically (by clicking on-screen), or manually (by typing in known values). In addition, it is also possible to use automatically-derived values, determined uniquely for each pattern. These can follow a pre-set sequence, or be pre-calculated from a diffraction pattern analysis. This allows diffraction patterns to be automatically sliced into 'segments', or allows azimuthal integrations around any number of radial maxima.

Graphically setting integration limits by 'drag-and-drop'

The SMC application also offers batch-wise intensity correction and normalisation. This can be used to correct for variations in beam brilliance, scale diffraction patterns to phase-specific contributions or adjust data sets for a changing sample volume. The software can automatically handle data organisation, records data treatment protocols and includes standard features such as a notepad and saved settings.

Before executing lengthy data reduction tasks, it is possible to preview the expected output for single files based upon the current program settings, as shown below.

The POI utility is designed to automate the analysis of large data series through linear and non-linear function fitting, and the batch-wise determination of other (more specialised) parameters. The program incorporates many special features which help to ensure that the results produced are reliable whilst overcoming the problems associated with diffraction data. This is essential what sets this software apart from the myriad of other fitting programs available. Performing batch-wise data analysis involves the straightforward selection of multiple scattering profiles from a file list, as shown below:

As the entire operation is GUI-based there are no complex commands to learn. Although processing rates depend upon the options chosen and computer specifications, 60 ms per file (including disk operations and display) are reasonable. This makes the batch-wise analysis of even large data series relatively quick.

The POI software offers a range of standard and specialised peak functions, as well as various linear and non-linear background functions. Multiple functions are also fully supported. Fitting is carried out using a modified Levenburg-Marquardt algorithm. Initial fit parameter estimates can be graphically set from a loaded data file. The fitting engine includes peak tracking, spike removal and file filtering as standard. This latter option can be used to skip low-intensity scattering profiles and therefore reduce the processing time wasted on unreliable (or unwanted) data.

Other specialised features include a function elimination algorithm and spread function fitting. Function elimination can be used to fit data series in which the number of peaks may change. Thus, the POI application can handle phase-changes and local texturing effects without manual data segregation. The spread function fitting option provides a robust method of fitting azimuthal profiles in which the peak position can vary around all azimuthal degrees. This can be coupled with specialist functions which also account for 'wrap-around intensity', incorporating asymmetry effects, whilst still sharing common fit parameters.

Finally, in addition to linear and non-linear function fitting, the POI utility can also calculate several other parameters on-the-fly. These include orientation functions and integrated intensities.

Effective result visualisation is one of the most critical aspects of treating large data sets. Not only is it important in terms of designing an analysis strategy, but it is also important in terms of result presentation. Furthermore, there is no advantage in an automated data reduction and analysis route if each result file has to be manually reformatted for use with a specific plotting program!

In terms of diffraction pattern visualisation, specialist software has been developed which can rapidly review large diffraction pattern series. Even at full detector resolution, images can be loaded and displayed in 200 ms or less, allowing whole experiments to be replayed like a movie! Indeed, just like a DVD it offers on-the-fly scaling, zooming and panning. Furthermore, coupled with an ability to show 1D scattering profiles and automatic background subtraction, no diffraction pattern feature can possibly be missed.

Other visualisation software developments have concentrated on analysis results, rather than the diffraction patterns themselves. This includes 2D area plotting of any analysis parameter according to the real-space scan area (allowing contour plots, vector plots, advanced filtering, overplotting etc...). A full statistical analysis package can instantly show parameter frequencies, advanced cross-correlations and a host of other statistical information at the click of a button. Meanwhile the treatment of 1D series includes normalisation, offseting, formatting, masking and cropping features which are fully compatible with any spreadsheet (allowing direct pasting). More detials will be provided on all of these software developments in due course.