NEWS

User-visible changes to festival-freebsoft-utils:

* Changes in festival-freebsoft-utils 0.10

** Speech Dispatcher support updated for new libc/iconv.

Due to a change in iconv Speech Dispatcher support has stopped to work.  This
release fixes the problem.

** TMPDIR environment variable honored.

There is a bug in the current released version of Edinburgh Speech Tools
resulting in ignoring temporary directory setting in TMPDIR environment
variable.  festival-freebsoft-utils now contains workaround of this bug.

** Bug fixes.


* Changes in festival-freebsoft-utils 0.9

** Character recoding improvements.

Recoding of characters from UTF-8 to 8-bit codings is more kind to extra
characters, not present in the target coding.  New variable
`recode-special-utf8-translations' allows to define custom mapping of extra
characters.

** New function `string-replace'.

** Bug fixes.


* Changes in festival-freebsoft-utils 0.8

** New features required by Speech Dispatcher 0.6.4.

** Additional language codes for available Festival voices.

** Bug fixes in SSML support.


* Changes in festival-freebsoft-utils 0.7

** Better support of the general intonation method.

It is generally possible to change basic prosodic parameters of voices that use
general intonation method.

** New languages supported.

Bug fixes and improvements that make Finnish, French and Italian voices work
with festival-freebsoft-utils.

** New utility functions `butlast' and `abs'.

** Use iconv instead of recode.

Iconv is more stable and better maintained recoding utility than recode.


* Changes in festival-freebsoft-utils 0.6

** Manual licensed under both GFDL and GPL.


* Changes in festival-freebsoft-utils 0.5

** SSML mode rewritten.

Major cleanup of the mode.
SSML interaction with Speech Dispatcher improved.
Additionally spelling support was added, while SSML prosody settings remained
unimplemented on the synthesis level in this version.

** New utility functions `min' and `max'.


* Changes in festival-freebsoft-utils 0.4

No user visible changes, just bug fixes.


* Changes in festival-freebsoft-utils 0.3

** New module fileio.scm.

** New module recode.scm.

** Changes of the Speech Dispatcher interface.

The Speech Dispatcher interface was changed to be compatible with Speech
Dispatcher 0.5 and is not compatible with older Speech Dispatcher versions.


* Changes in festival-freebsoft-utils 0.2

** Speech Dispatcher rate and pitch setting changed.

Now the Speech Dispatcher to Festival pitch/rate value mapping is smooth and
pitch is set relatively to the default pitch of current voice.

** Event handling mechanism was reworked.

Event handling was redesigned to produce better results and to be cleaner and
more robust.

** Voice and language selection mechanism was generalized and simplified.

There are new language-codes and voice-select-defaults variables that replace
the old speechd-* voice and language selection variables.

** New functions for unified setting of prosodic parameters.

With their help, you can change basic prosodic parameters like pitch, rate or
volume easily.

** There is new option for disabling initial pauses in synthesized samples.

See the variable inhibit-initial-pauses.

** Utterance chunking implemented in SIOD.

It is now possible to perform utterance chunking not only in tts_file, but in
any extension function.

** Synthesized events can be returned in multiple wave forms.

See the manual entry of the multiple.scm module for description why this can be
useful.  The Speech Dispatcher interface was enhanced this way too.

** Rudimentary SSML support.

** New utility functions.

See documentation of the util.scm and wave.scm files for the complete list of
available utility functions.

** SoX is required.

The SoX sound processing tool is now required for most of the functionality.


Local variables:
mode: outline
end: