Speech recognition typically works with large corpora. Therefore almost all programs in SPRAAK are written from the viewpoint of processing multiple files constituting a corpus. Along the same lines we work more often with scripts that execute a set of routines.

Of course it is possible to execute single commands on single files as well as these are always the building blocks of the corpus based scripts.

Some of the commands that are most handy for inspecting data and debugging purposes are:

spr_copy.c Copying and Printing of SPRAAK files
spr_sigp.c Master Signal Processing Routine
spr_vitalign.c Viterbi Alignment
spr_cwr_main.c The SPRAAK continuous speech recognizer