HOWTO: Make festival TTS use better voices (MBROLA / CMU / HTS)

[chovy2]

MBROLA instructions don't work anymore

[wayward4now]

MBROLA instructions don't work anymore

I would really like to see a workup fixing that
"Error: HTS_Model_load_pdf: Failed to load header of pdfs." error resolved. That last thing I want is "man-talk" by my computer! <grins> Ric

[wayward4now]

p/s Thanks for sticking with this thread for five years! I'm impressed! Ric

[wayward4now]

I'm blubbering a heartfelt THANKS!! It works! She is not the Honey West I had hoped for, but she sounds better than Madonna.

[apurva2]

hey guyzz plz help me ...in my /usr/share/festival/voices/english....i hv kal_diphone folder which is english male voice bt nw i want to change english male voice to female voice so ..i downloded " cmu_us_clb_arctic-0.95-release.tar.gz ...then followed following step ...
cd /usr/share/festival/voices/english/
sudo tar jxf cmu_us_clb_arctic-0.95-release.tar.bz2
sudo ln -s cmu_us_clb_arctic cmu_us_clb_arctic_clunits
sudo cp /etc/festival.scm /etc/festival.scm.backup
sudo echo "(set! voice_default 'voice_cmu_us_clb_arctic_clunits)" >> /etc/festival.scm

error is bash : /etc/festival.scm : permission denied...
so i used following cmd =
gksu gedit /etc/festival.scm

after that /etc/festival.scm file is opened and then at the end of file i typed (set! voice_default 'voice_cmu_us_clb_arctic_clunits) and then i saved it
after that i typed festival on terminal it has given following error :
SIOD ERROR: unbound variable : f2b_f0_lr_start
closing a file left open: /usr/share/festival/init.scm
festival: fatal error exiting.


so i again opened /etc/festival.scm and removed that
(set! voice_default 'voice_cmu_us_clb_arctic_clunits)
and then it strted to work...so plz tell me how to change voice from male to female...

[Graham_Toal]

There's a commented-out line in /usr/share/festival/voices/english/cmu_us_awb_arctic_clunits/festvox/cmu_us_awb_arctic_f0model.scm:
;(require 'f2bf0lr)
just add it back in...

[vahnskeyblade]

Hi there,

I followed your instruction verbatim, and for the most part, they were successful. Once I get to actually running it, though, and I've set one of the nitech hts voices as my default, this is what I get in response:

Warning: HTS_fopen: Cannot open hts/htsvoice.
aplay: main:593: bad speed value 0
aplay: main:593: bad speed value 0
nil

So, I don't know what to do from here. Hopefully, this is not a hard problem to fix. I am very new to linux and ubuntu, and am teachable as far as the terminal goes, but I don't have a lot of understanding, yet. Most of my learning has been through Google.

[hiddenotebook]

Lets revive this great post guys I finally can install cmu voices on ubuntu 16.04 check this how to:

Make a directory to do the job:

mkdir myvoice  
cd myvoice

Download this great and clear voice:

wget http://www.speech.cs.cmu.edu/cmu_arctic/packed/cmu_us_slt_arctic-0.95-release.tar.bz2

Extract it:

tar xf cmu_us_slt_arctic-0.95-release.tar.bz2

Now move it to festival. Beware, we need to change the name here add _clunits suffix ok:

sudo mv cmu_us_slt_arctic /usr/share/festival/voices/english/cmu_us_slt_arctic_clunits

Make it the default voice : edit festival config file add this to a new line (/etc/festival*.scm)

(set! voice_default 'voice_cmu_us_slt_arctic_clunits)

Now start festival:

festival

Test it:

(SayText "Hello I am Alice. hot and nice. This is a short introduction test.")

Works in Ubuntu 16.04 and raspbian jessie

[frytek]

This post has nothing to do with festival. However, I'm using this thread to ask the following question: is anybody interested in trying out a virtual machine with Lubuntu and a preinstalled speech synthesis system using the Google OpenSAPI engine? If so send me a private message.

[adzcsx2]

Thank you so much! You have saved me . I am registering for you .