The package contains two programs - Waver and Train_verif. The programs are dedicated for processing the Corpora - Polish speech data basis designed by Stefan Grocholewski.
The first one was designed for joining files of this data basis and the second one - for spliting the joined files into two parts: training and veryfying.
The package contains both delphi sources and executable form.
For convenience prepare directory structure:

apl\waver 				- place here delphi sources and executable form
apl\waver\data 				- place here list of the Corpora's files names
apl\waver\results			- direct here results of processing
apl\waver\description 			- place here description of the program

and 

apl\waver\train_verif_samples 			- place here delphi sources and executable form
apl\waver\train_verif_samples\results 		- direct here results of processing
apl\waver\train_verif_samples\description 	- place here description of the program

To obtain the Corpora data basis turn to http://libra.cs.put.poznan.pl/research/labOper/index.html