EDINBURGH SPEECH TOOLS LIBRARY
                 VERSION 1.2.96 August 2006

Principal authors: Alan W Black, Paul Taylor, Richard Caley, 
                   Simon King, Rob Clark and Korin Richmond

Centre for Speech Technology
University of Edinburgh 
2 Buccleuch Place
Edinburgh EH8 9LW, UK
http://www.cstr.ed.ac.uk/projects/speech_tools

Although this is a considered stable release, there are still a number
of known parts are missing or incomplete.

------------------------------------------------------------------------

The Edinburgh speech tools system is a library of C++ classes, functions
and utility programs that are frequently used in speech software. The
system compiles to a single unix library .a file which can be linked
with software. At present, C++ classes for several useful speech and
language classes have been written, along with audio software and some
basic signal processing software.

========================================================================


            Edinburgh Speech Tools Library version 1.2.96
              Centre for Speech Technology Research                
                   University of Edinburgh, UK                     
                    Copyright (c) 1994-2006
                      All Rights Reserved.                         
                                                                   
Permission is hereby granted, free of charge, to use and distribute  
this software and its documentation without restriction, including   
without limitation the rights to use, copy, modify, merge, publish,  
distribute, sublicense, and/or sell copies of this work, and to      
permit persons to whom this work is furnished to do so, subject to   
the following conditions:                                            
 1. The code must retain the above copyright notice, this list of    
    conditions and the following disclaimer.                         
 2. Any modifications must be clearly marked as such.                
 3. Original authors' names are not deleted.                         
 4. The authors' names are not used to endorse or promote products   
    derived from this software without specific prior written        
    permission.                                                      
                                                                     
THE UNIVERSITY OF EDINBURGH AND THE CONTRIBUTORS TO THIS WORK        
DISCLAIM ALL WARRANTIES WITH REGARD TO THIS SOFTWARE, INCLUDING      
ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS, IN NO EVENT   
SHALL THE UNIVERSITY OF EDINBURGH NOR THE CONTRIBUTORS BE LIABLE     
FOR ANY SPECIAL, INDIRECT OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES    
WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN   
AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION,          
ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF       
THIS SOFTWARE.                                                       

========================================================================

The following c++ programs are available:

	na_play: generic playback program for use with net_audio and CSTR ao.

	ch_wave: Waveform file conversion program.

	ch_lab: label file conversion program.

	ch_track: Track file conversion program.

        wagon: a CART tree build and test program

        And others

The following C++ sub-libraries are available

	audio:	C++ audio functions for Network Audio system, Suns
                Linux and FreeBSD

	speech_class:	C++ speech classes, including waveform and track.

	ling_class:	C++ linguistic classes.
	
	sigpr:	Signal processing

	utils: Various utilities.	


COPYING
------------------------------------------------------------------------

Some files in the distribution do not fall under the above copyright.
However none have any commercial restrictions, and no files fall under
the GPL.  Specifically files that have different licences are as
follows:

grammar/wfst/wfst_train.cc
        Copyright Carnegie Mellon University (under the same
        open licences as much of the rest of the code.
siod/ 
include/siod*.h
lib/siod/siod.scm
        Apart from some new functionality under our licence the SIOD
        files fall under the licence given in siod/slib.cc copyright
        Paradigm Associates.  This code has been substantially modified
        by us though our changes are under that licence.

        editline.c editline.h el_complete.c el_sysunix.c el_unix.h
        are copyright 1992 Simmule Turner and Rich Salz under the
        "BSD like" licence in siod/editline.c  Again substantial 
        changes have been made by us which continue to be under that
        licence

rxp/    
include/rxp/
        Is Copyright Richard Tobin.  We have a specific version of RXP
        which allows free (including non-commercial) use.

base_class/raveconv.cc
        Is copyright 1992, 1995 by Markus Mummert

base_class/string/reg*.{c,cc}
        Copyright (c) 1986 by University of Toronto.

base_class/string/EST_strcasecmp.c
   Copyright (c) 1987, 1993
 	The Regents of the University of California.  All rights reserved.

   The standard BSD licence


ACKNOWLEDGEMENTS
------------------------------------------------------------------------

Most of the software was written by Paul Taylor, Alan W Black, Simon
King, Richard Caley, Rob Clark and Korin Richmond. Additional
contributions have been made by the following people.  Slightly
different licence agreements accompany these pieces of code, so it is
important that the various agreements be properly understood.

Tony Robinson: some signal processing code.
Markus Mummert: sample rate converter (rateconv.cc).
Craig Reese and Joe Campbell: ulaw conversion code.
Paul Bagshaw: pitch tracker.
Theo Veenker: IRIX audio support.
Stan Chen: AIX support
Paradigm Assoc. and George Carrett: For Scheme In One Defun
Simmule Turner and Rich Salz: for editline
Henry Spencer: for regex code
The Regents of the University of California: some string comparison code
Richard Tobin: RXP, XML parser.
Arthur Toth: Windows server code

INSTALLATION
************

see INSTALL file for details

FILE/DIRECTORY STRUCTURE
------------------------------------------------------------------------

The main directory contains minimal wrap-around main() functions
which compile to form executables which use the algorithms in the
speech tools. All these main() functions really do is to read
command line options and read and write files, the algorithms
themselves are completely contained within the libraries.

The directory structure is as follows:

config: 
        shared make rules and configuration make rules

speech_class:
	general speech class files, including waveforms, contours etc.

ling_class:
	general linguistic class files, including a stream-based architecture.

sigpr:
	signal processing code.

doc:
	basic documentation, generates html, info and postscript

include:
	class and general header files.

lib:	
	where the .a files are kept

main:
	Directory containing main files which compile to executables.


DOCUMENTATION
------------------------------------------------------------------------

There are two sources of documentation.

Man pages exist for the more important stand-alone programs. These are
found in the main directory.

The library documentation has substantiall improved in this version
using an integrated DOC++, JADE and docbook combination. Classes are
described at the overview, theorectical, practical and code level.
The build process for the documentation is still dependent on some
more obscure packages so it is recommended that you take the
pre-built documentation.  Or access it on-line through
   http://www.cstr.ed.ac.uk/projects/speech_tools

The library documentation is in texinfo format which can be converted
into GNU info HTML and postscript, see Makefile in doc/

Note the documentation is not up to date. We are in the process of moving the documentation to Doxygen.

BUGS/COMMENTS ETC
------------------------------------------------------------------------

Substantial parts of this release are considered mature and should be
relatively free of bugs, however there are newer parts which have not
been rigourously tested yet.  If you find any bugs, please send a bug
report to speechtools@cstr.ed.ac.uk. Any useful comments will also be
appreciated. See also the bugs file in the doc directory.

If you have some software that you think could be integrated into the
speechtools system, please contact us at
speechtools@cstr.ed.ac.uk. The whole point of writing and releasing
code such as this is to prevent duplication of effort, and therefore
we would appreciate greatly any code which will add to the usefulness
of the system. Signal processing routines are particularly welcome.


======================================================================
ADDITIONAL SOFTWARE
======================================================================

CSTR supplies a number of different speech technology systems
including speech synthesis systems and (in the future) a recognition
system.  To find out more information checkout http://www.cstr.ed.ac.uk/