Figure 1
The dxtbx data model for a complex set of input images. The image files (1) are fed into the data block factory (2). The data block factory then uses the format registry (3) to interrogate each image to find the Format class (4) that best understands it. Note that the dxtbx supports more Format classes than are shown in the figure. If all the images use the same Format class, then a single data block (5) is returned; otherwise, multiple data blocks are created. The data block analyses the image metadata to group the images together on the basis of the set of experimental models that are shared between them. These groups can be accessed from the datablock as either sweeps or sets (6). A sweep must contain a beam, detector, goniometer and scan (7), and is thus appropriate for rotation photography; an imageset must have a beam and detector model for each image, as for a set of still shots. |