Page 14 of 19 FirstFirst ... 41213141516 ... LastLast
Results 131 to 140 of 181

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

  1. #131
    Join Date
    Dec 2010
    Beans
    Hidden!
    Distro
    Ubuntu 10.04 Lucid Lynx

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

    We will develop an additional festvox voice. Therefore I had to test the Festival installation process. So far the Nitech voices are the best I came across.
    Thanks for your helpful tutorial

    Quote Originally Posted by mrplow View Post
    ...I've spent enough time and I'll wait until someone tries out that new festival and reports back
    hi mrplow. I compiled Festival 2.1. successfully yesterday evening following the detailed instructions of every README and INSTALL text.

    But I had problems getting the Nitech voices to work due to "failed to load header of pdfs". [1] I didn't really understand the proposed solution so I was not able to solve the problem. I completely removed the files today and tried with the old repository version which worked like a charm.

    [1] http://hts.sp.nitech.ac.jp/hts-users.../msg00196.html
    Last edited by meufel; January 14th, 2011 at 10:04 PM.

  2. #132
    Join Date
    Dec 2010
    Beans
    Hidden!
    Distro
    Ubuntu 10.04 Lucid Lynx

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

    sorry doublepost
    Last edited by meufel; January 15th, 2011 at 01:08 PM. Reason: doublepost

  3. #133
    Join Date
    Apr 2009
    Beans
    1

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

    Does anybody know how to use multiple -eval paramaters in the text2wave program by Festival?

    I would like to set (voice_cmu_us_awb_arctic_clunits) as well as (Parameter.set 'Duration_Stretch 1.5) but I cant figure out how?

    Any help?

  4. #134
    Join Date
    Feb 2011
    Beans
    3

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

    Quote Originally Posted by meufel View Post
    We will develop an additional festvox voice. Therefore I had to test the Festival installation process. So far the Nitech voices are the best I came across.
    Thanks for your helpful tutorial



    hi mrplow. I compiled Festival 2.1. successfully yesterday evening following the detailed instructions of every README and INSTALL text.

    But I had problems getting the Nitech voices to work due to "failed to load header of pdfs". [1] I didn't really understand the proposed solution so I was not able to solve the problem. I completely removed the files today and tried with the old repository version which worked like a charm.

    [1] http://hts.sp.nitech.ac.jp/hts-users.../msg00196.html
    hi Ubuntu/Festival GURUs

    I am new to Ubuntu/Linux and new to festival i am trying to setup festival with Nitech voices but you all already struggled to setup those with latest version of festival 2.1. So I also decided to use festival 1.96.

    for this I struggled to use apt-get install to install festival-1.96 but i don't know how to install previous version where as the current version is now 2.0.96~beta, if someone please help me in this considering me a dump on linux.


    Anyway I moved to the instruction on Post#1 to build festival1.96~beta myself but when i try to debuild it (after performing all the prerequisites) i got following error on terminal.

    Please someone help me to get this fixed. Thanks before hand

    Code:
    EST_wave_io.cc:74: error: invalid conversion from ‘const char*’ to ‘char*’
    EST_wave_io.cc: In function ‘char* nist_get_param_str(const char*, const char*, const char*)’:
    EST_wave_io.cc:90: error: invalid conversion from ‘const char*’ to ‘char*’
    EST_wave_io.cc: In function ‘EST_read_status load_wave_nist(EST_TokenStream&, short int**, int*, int*, int*, int*, EST_sample_type_t*, int*, int, int)’:
    EST_wave_io.cc:222: warning: ignoring return value of ‘int system(const char*)’, declared with attribute warn_unused_result
    make[2]: *** [EST_wave_io.o] Error 1
    make[1]: *** [speech_class] Error 2
    make[1]: Leaving directory `/home/mubashar/fest_tmp/speech-tools-1.2.96~beta'
    make: *** [build-stamp] Error 2
    debuild: fatal error at line 1319:
    couldn't exec fakeroot debian/rules:

  5. #135
    Join Date
    Feb 2011
    Beans
    3

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

    I am new to Festival and New to Unix/Linux envoirnment.

    I have Setup Festival 2.1 on Ubuntu 10.10 successfully with instruction given in INSTALL file.

    Then I decided to setup HTS 2.1.1 voices available at http://hts.sp.nitech.ac.jp/?Download.

    Downloaded, Configure, make and make install its prerequisites and HTS-demo_CMU-ARCTIC-SLT as per instruction given in INSTALL file and this post

    Everything gone smooth but what to do now, should I copy some files to festival folder, do i need to tell the path of HTS-demo_CMU-ARCTIC-SLT to festival somehow or to add it to Some Environment Variable.

    Please answer me with consideration that I am new to Unix/Linux environment If some one can share the detailed guidelines, I would also like to add other avialble voices in HTS 2.1.1.

  6. #136
    Join Date
    Sep 2007
    Beans
    2

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

    A workaround:

    * Install VirtualBox
    * Install Ubuntu 8.04.4 LTS
    * Follow the instructions in the original post. It works perfectly.

  7. #137
    Join Date
    Dec 2009
    Beans
    32
    Distro
    Ubuntu 9.10 Karmic Koala

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

    Sounds to me like I'll have to wait for festvox to release the HTS 2.1.1 voices if I want to use festival 2.0.95 beta? Or I could figure out how to compile the voices myself... Decisions, decisions. Also, the previous version of festival in Ubuntu's repos (<= Lucid, 1.96 beta) is from 2004 and 2.0.95 beta (>= Maverick's repos) is from 2010. So should I...

    a) Use an outdated version of festival with better voices
    b) Use a newer version of festival with poorer quality voices
    c) Figure out how to build the voices from source

    Sounds like there are enough of interested in it, I'll post back with any progress I make. Please do the same

  8. #138
    Join Date
    Feb 2011
    Beans
    3

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

    Quote Originally Posted by LCamel View Post
    A workaround:

    * Install VirtualBox
    * Install Ubuntu 8.04.4 LTS
    * Follow the instructions in the original post. It works perfectly.
    Thanks for Reply, It really WORKED!!!!

    Some other voices are not working but nitech voices are working quit fine, anyway I am looking for to control the speed of utterance is there anyway to do it????

  9. #139
    Join Date
    Feb 2011
    Beans
    4

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

    Hello, I want to use this scripts for training my own voices, and I have this problem:

    Synthesizing a speech waveform from /home/gtc/dmolina/dmolina/hts_pruebas/ar2_michel/data/labels/gen/michelgen1.lab using hts_engine...

    Error: HTS_Model_load_pdf: Failed to load header of pdfs

    I would like to know how can I solve it, because I need to use this voices

    Thanks

  10. #140
    Join Date
    Jun 2011
    Beans
    2

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

    SIOD ERROR: could not open file /usr/lib/festival/siod.scm
    SIOD ERROR: could not open file /usr/lib/festival/siod.scm


    I install festival, speech_tools, festvox (everything i need it) but i don't know what to do in this case. Someone have any idea about it?

    thanks



Page 14 of 19 FirstFirst ... 41213141516 ... LastLast

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •