eSpeak 1.16   02.Pct.2006

NOTE:  The contents of the files in the espeak-dat
  directory, and the format of their contents has changed
  for this version. Please delete your old  espeak-dat
  directory and replace it with the new one in this
  package.

1. espeakedit now includes a makefile.

2. Added dental variants of [t] and [d] phonemes for
   it, es, ru, pl

3. Added palatal [l^] for ru, it ("gli").

4. Fixed a bug with portaudio interface aborting speech
   in some circumstances.


eSpeak 1.15   25.Sep.2006

1. Added "length" field to espeak_EVENT structure in
   speak_lib.h.


eSpeak 1.14   21.Sep.2006

1. Makefile now produces both the "speak" program and
   the shared library version:  libespeak.so.1.1.14

2. Most languages now speak groups of digits rather than
   just single digits.
   Translating groups of digits into spoken numbers is now
   done by a program function: TranslateNumber() rather
   than by a large number of rules in the xx_rules file.
   The number fragments used by TranslateNumber() are
   found in the xx_list file and their labels all begin
   with underscore followed by a digit.

3. Improvements to Italian, German.
   Initial version of Welsh and Russian (note, a larger
   Russian dictionary file is available as a separate
   download).

4. Added functions to libespeak API (see speak_lib.h):
   void espeak_SetPhonemes(int value, FILE *stream);
   void espeak_CompileDictionary(const char *path, FILE *log);

5. Added some additional experimental intonation types.
   To try these add a line to the end of a voices file:
      intonation <type>

   0 is the default, the other values are 1, 2, 3, 16.

   Let me know if you want to experiment with writing a
   different intonation algorithm.



eSpeak 1.13   30.Aug.2006
1. Now compiles for use with either V18 or V19 of
   PortAudio.  Copy either portaudio18.h or portaudio19.h
   to portaudio.h.

2. Can be used as a library rather than a separate
   process. Compile with source file  speak_lib.cpp
   rather than speak.cpp.
   The API is described in  speak_lib.h.

3. Added language: Greek (just guesswork, it needs work by
   a native speaker).

4. Various improvements, including the Italian voice.


eSpeak 1.12   29.Aug.2006
   Withdrawn due to an error.


eSpeak 1.11   07.Aug.2006

1. Recognise and obey some SSML (Speak synthesis markup
   language) and HTML tags, and ignore unrecognised tags.
   Use with the -m option.

2. Added language: Afrikaans.

3. Added initial draft implementation of languages which
   need work and advice from native speakers:
    Spanish, Italian, Polish,
   as well as previous: German, Esperanto.

4. English languages variants (Voice file now renamed to:
   en-n, en-rp, en-wm) are improved.

5. Added --voices command option to list voices.  Voice
   files can now be in subdirectories in
   espeak-data/voices.



eSpeak 1.10   29.Apr.2006

1. Add UTF-8 capability.  Assume input text is UTF-8 until
   any non-UTF-8 sequences are found.

2. Use UTF-8 encoding for eSpeaks's data files (_rules,
   _list, _extra)

3. Treat any accented characters which are not recognized
   in the _rules data as their un-accented equivalents.

4. Improve pronunciation of number strings.

5. Added embedded commands within the input text to change,
   pitch, speed, amplitude, etc.  See docs/ReadMe for details.

6. Added command option  --punct  to speak punctuation within
   the input text.

7. Added command option  -k  to indicate capital letters
   within the input text by either a click, raised pitch,
   or spoken word "capital".

8. Changed command options for phoneme display from -p -P
   to -x -X.  Added new option -p to adjust pitch.

9. Improvements to Esperanto language, including numbers.

10. Added "roughness" attribute for voice files.



eSpeak 1.09   10.Apr.2006

1. Changes for use with ARM/RISCOS version.

2. Added "flutter" attribute for voice files.



eSpeak 1.08   06.Apr.2006

1. Added command syntax to speak text string:
      speak [options] "this is some text"

2. Added command option --stdout to write speech output to
   stdout

3. Change SpeakNextClause() function to take input from
   either  a file or memory.

4. Changed compiler options in Makefile

5. Added phonemes @- Q  o- u-

