Linux Festival text-to-speech .mp3 production instructions.

Linux Users: produce text-to-speech .mp3 podcasts digitally in much less than realtime, in two steps.

in bash:

$ text2wave -otype wav -F 44100 (yourfilename).txt -o (yourfilename).wav

$ lame -fb 96 (yourfilename).wav (yourfilename).mp3

(no parenthesis; "( )")

dependencies:

Festival
Festivox (voice)
Lame

Helpful:

KTTS
Kmouth
Ksayit
Audacity

Special thanks to my new friend, mike@hingston.demon.co.uk

<a href>http://www.hingston.demon.co.uk/mike/linuxtravels.html#festival2/<,/a>

Batch Process:

I do a daily podcast Bible Study and post once a week. Linux KTTS allows pronunciation correction, but only one day at a time, in real time. I use a perl looping script and a Bash script; the first to find and replace Biblical words with phoenitic representations (Malachi = Mal ah kai). This is an incredible time-saver, compared to real-time analog tts recording. Words can be constantly added to this file as needed.

I currently post mp3 files to:

www.archive.org/details/audio

see links at:

http://shepboy.snow.prohosting.com/dw_bible2/B%20Year/wklx_b.html

The perl looping command is:

perl "s/search_text/replace_text/g" filename

example:

perl -w -i -p - e- "s/Malachi/Mal ah kai/gt" *.txt

This can be repeated as many times as necessary to modify every word you wish to change.Place script in folder to be executed. "cd" to folder; Don't execute globally.

This will search and replace each word in each file of the .txt extension in the folder.

example:

===
#/bin/bash
perl -w -i -p -e 's/Malachi/Mal ah kai/g;' *.txt
perl -w -i -p -e 's/Samaria/Sa mari a/g;' *.txt
perl -w -i -p -e 's/Hoshea/Ho she a/g;' *.txt
perl -w -i -p -e 's/Shalmanesar/Shalman e sar/g;' *.txt
perl -w -i -p -e 's/Ahab/A hab/g;' *.txt
perl -w -i -p -e 's/II/2/g;' *.txt
perl -w -i -p -e 's/Caesarea/Ses ar ia/g;' *.txt
perl -w -i -p -e 's/Shechaniah/Shech a niah/g;' *.txt
perl -w -i -p -e 's/Aramean/A ra me an/g;' *.txt
perl -w -i -p -e 's/Maranatha/Ma ra na tha/g;' *.txt
perl -w -i -p -e 's/Nebuchadrezzar/Neb u kad rezzar/g;' *.txt
perl -w -i -p -e 's/Nebuchadnezzar/Neb u kad nezzar/g;' *.txt
perl -w -i -p -e 's/Artaxerxes/Art ax erses/g;' *.txt
perl --w -i -p -e 's/Ahava/Ahav a/g;' *.txt
perl -w -i -p -e 's/Eleazar/El e a zar/g;' *.txt
perl -w -i -p -e 's/Capernaum/Ca pern ne um/g;' *.txt
perl -w -i -p -e 's/Sihon/sy on/g;' *.txt
perl -w -i -p -e 's/Meremoth/mer emoth/g;' *.txt
perl -w -i -p -e 's/Antipas/Anti pas/g;' *.txt
perl -w -i -p -e 's/herodias/Har odias/g;' *.txt
perl -w -i -p -e 's/Ahaz/Ayhaz/g;' *.txt
perl -w -i -p -e 's/Ananias/An an i as/g;' *.txt
perl -w -i -p -e 's/Archelaus/ark e lay us/g;' *.txt
perl -w -i -p -e 's/Immanuel/Im man u el/g;' *.txt

#more lines, as needed, follow here to eof
#eof
#The options forms a memorable acronym.
===

The basic text to .mp3-multi:

Place script in folder with multiple files with same extension *.txt to be converted. "cd" to folder; Don't execute globally!

Note: when this first executes, Bash will report that .wav file cannot be opened. This is normal. Just wait a few minutes and the second part of the script will execute, with visible progress.

Make sure that all files in folder have the same name syntax, and the same *.txt extension; suspect syntax error for any files not converted.

The following script assumes a (unix) ".txt" file for each day of the week; otherwise you will have to modify. Check metadata properties.
===
#!/bin/bash

if test -z "$1"
then
clear
echo 'You must enter a File to convert'
echo 'EX: 1advent_a'
echo '...where input is 1advent_a_sun.txt'.
echo 'NOTE: NO FILE EXTENSION {_day.txt assumed}'
exit
else
clear
      /usr/bin/text2wave -otype wav -F 44100 $1_sun.txt -o $1_sun.wav
     /usr/bin/lame -fb 96 $1_sun.wav $1_sun.mp3

    /usr/bin/text2wave -otype wav -F 44100 $1_mon.txt -o $1_mon.wav
     /usr/bin/lame -fb 96 $1_mon.wav $1_mon.mp3

     /usr/bin/text2wave -otype wav -F 44100 $1_tues.txt -o $1_tues.wav
     /usr/bin/lame -fb 96 $1_tues.wav $1_tues.mp3

     /usr/bin/text2wave -otype wav -F 44100 $1_wed.txt -o $1_wed.wav
     /usr/bin/lame -fb 96 $1_wed.wav $1_wed.mp3

    /usr/bin/text2wave -otype wav -F 44100 $1_thurs.txt -o $1_thurs.wav
     /usr/bin/lame -fb 96 $1_thurs.wav $1_thurs.mp3

     //usr/bin/text2wave -otype wav -F 44100 $1_fri.txt -o $1_fri.wav
     /usr/bin/lame -fb 96 $1_fri.wav $1_fri.mp3

    /usr/bin/text2wave -otype wav -F 44100 $1_sat.txt -o $1_sat.wav
     /usr/bin/lame -fb 96 $1_sat.wav $1_sat.mp3
fi
exit

===
<a href >http://crazytoon.com/2007/10/29/linux-how-do-i-mass-find-and-replace-text-in-files-under-linux-using-perl/</a>