
and i hereby offer it for everyone to use, too. The Process is completely automated at this point, you just have to follow a few rules in your t-file. The only "downside" is that i am very unfamiliar with the Windows shell, so the package is a linux script and some xsl files only - but i will generate all Files for you if you provide me the t-file so dont worry about that

and now first things first: rules for the t-file:
A) the t-files must be merge files, not diff files, meaning they have the same structure as the original t-files and simply use new page ids because my script cannot work with diff files (and it is not necesary to use them imo since you usually only add entries)
B) for the Text to be Voiced it is better to use new Pages and dont add to existing ones to not confuse the Game
C) in each Page to be voiced set the » voice="yes" « Attribute and for completeness also specify the TTS Engine and the Options you want to use.
currently supported are
-> pico2wave (default & similiar to the original BC voice) => add a » pico2wave-options «-Attribute to the Page Node
-> espeak => Add a » espeak-options «-Attribute to the Page-Node
the Options of these Engines can be set here, too (either a default per Text Page or you can also specify Settings per t-id with the same attribute

D) do not use "" or '' in your Text because these will be ignored by my Script - or alternatively use the » speak="Text to Speak" « Attribute in the t node, which defines an alternate Text to be spoken (also useful if you are not satisfied with the Output and want to refine the Text

these adittional Attributes wont give any Errors, so simply leave them there

I also take suggestions for more TTS Engines to add, only condition is they are useable from the Command Line in Linux

Possible Options (default as example):
pico2wave:
--lang=en-US
=> sets the Language for this TTS-Engine, also available is de-DE for example
espeak: too many to list here, see here under point 2.2.2 The Command Line Options
most useful:
-v en-1
=> sets the voice/language. english male voices are numbered (iirc up to en-5) and female voices have an f before the number (up to en-f5 iirc, too)
german voices use de-5 for example, or de-f3
Example t Files: Roman Numerals from 1 to 5:
Code: Select all
<?xml version="1.0" encoding="UTF-8" ?>
<language id="44">
<page id="12345" title="Example: Roman Numerals" descr="s" voice="yes" pico2wave-options="">
<t id="1" speak="One">I</t>
<t id="2" >II</t>
<t id="3" >III</t>
<t id="4" speak="Four">IV</t>
<t id="5" speak="Five">V</t>
</page>
</language>
Code: Select all
<?xml version="1.0" encoding="UTF-8" ?>
<language id="49">
<page id="12345" title="Example: Roman Numerals" descr="s" voice="yes" pico2wave-options="--lang=de-DE">
<t id="1" speak="Eins">I</t>
<t id="2" >II</t>
<t id="3" >III</t>
<t id="4" >IV</t>
<t id="5" speak="Fünf">V</t>
</page>
</language>