Datasets are comprised of a DDS Dictionary (which defines the binary format and contains the processing history) and a Binary File. The Binary File can be attached to the Dictionary to form a single file for the Dataset. For predefined data formats like usp, segy, segy1, or su the Dictionary is optional, in which case certain default definitions are setup for these fixed predefined formats.
Filenames have a few constraints. They should not be terminated with ":" (colon). These names are reserved for internal mnemonic data streams. Filenames must not contain "=", white space, or control characters.