Start by making a Zip archive of your raw data. Leave the files exactly as they are coming off your instrument. CRAB will convert your data into TIFF images and JSON metadata automatically.
The currently installed version of CRAB supports data from the following devices:
CRAB organises most data into projects. Think of a project as a self-contained collection of data that all relates to a particular study. Each project can have several collections, with each collection being a different variation of the data. Many projects will only consist of the default "main" collection. Snapshots represent a collection at a particular point of time. Analagous to a research output, these snapshots hold all data and metadata frozen in time. CRAB export utilities are built around snapshots to support replicability. Someone else using your code, and your exact snapshot, should be able to replicate your findings and results.
CRAB presents many different ways to access your data. A common use case is exporting datasets to machine learning libraries such as Keras. To support this use case, CRAB generates Croissant files which can be accessed from a project's page after a snapshot has been taken.