Text-to-speech

Converted with HTML::WikiConverter::MediaWiki from my old phpwiki site

---

Installation de festival et MBROLA:

• http://leb.net/blinux/list-archive/blinux-list/1998/msg00956.html
• MBROLA
• http://linux.maruhn.com/cat/Tool/Sound.html

Download:

• wget http://tcts.fpms.ac.be/synthesis/mbrola/bin/pclinux/mbr301h.zip
  • unzip mbr301h.zip mbrola-linux-i386
  • mv mbrola-linux-i386 /usr/local/bin
• wget ftp://ftp.pbone.net/vol6/ftp.redhat.com/pub/contrib/libc5/i386/mbrola-ttp-English-1.0-4.i386.rpm
  • alien mbrola-ttp-English-1.0-4.i386.rpm
  • dpkg -i mbrola-ttp-English-1.0-4.i386.deb
• wget http://tcts.fpms.ac.be/synthesis/mbrola/dba/fr4/fr4-990521.zip
  • unzip
  • mkdir /usr/local/lib/mbrola/fr4

INSTALLATION

---

Here, I assume your linux sound hardware is already set up.

1. Install the following RPM packages for festival

festival-1.2.1-3.i386.rpm festdoc-1.2.0-1.i386.rpm festlex_OALD-1.2.1-1.i386.rpm festlex_POSLEX-1.2.1-1.i86.rpm festvox_en1-1.2.1-1.i386rpm

And this one for MBROLA support within festival

mbrola-ttp-English-1.0-3.i386.rpm

I got the above last September/October from a Blind Linux

(BLINUX) site. You can find more up-to-date versions from

http://leb.net/blinux/index.html

or ftp://leb.net/pub/blinux/

2. Install the MBROLA package. You'd probably prefer the RPMs from

BLINUX; however, I did mine manually from the .tar.gz files:

from mbr205b.zip

mbrola => /usr/local/bin/mbrola readme.txt => /usr/local/doc/mbrola-2.05b/readme.txt

from en1.zip

en1/* => /usr/local/lib/mbrola/en1/*

Next, create some symbolic links

cd /usr/lib/mbrola ln -s /usr/local/lib/mbrola/en1 cd /usr/lib/festival/lib/voices/english/rb_mbrola ln -s /usr/lib/mbrola/en1/en1 ln -s /usr/lib/mbrola/en1/en1mrpa

3. Insert a line in /usr/lib/festival/lib/voices.scm

(defvar default-voice-priority-list '(

rb_mbrola ; <= INSERT THIS LINE rab_diphone kd_diphone don_diphone gsw_diphone spanish_el )

4. You can test your setup by running festival like this

bash$ festival festival> (SayText[SayText?action=create ?] "Type any text here. Sounds cool?")

1. <UTT 0x1234567>

festival> (quit)

5. Festival can be configured to pre-process various file formats to

make them more "readable", e.g. HTML. See the festival documentation in /usr/doc/festdoc-1.2.0/festival/festival.info for information on how to set this up. It's under the section "Text modes".

As for me, I chose to write a good old sed script to do my text pre-processing instead.

6. You can run festival in text-to-speech mode on a file

festival --tts <filename>

or on a stream

pre-process <filename> | festival --tts

"pre-process" is the name of my sed-based text pre-processor.