Text to speech in Linga Franca Nova
   
         
  back   Text-to-speech (TTS) is the generation of synthesized speech from text.   www TCTS Lab.
          www google on tts
     
TSS has become a mature technology. For current languages like English, Japanese, Spanish, French or German, the produced speech is acceptable or good. To illustrate the state of the art, examples are included here. As this examples demonstrate, the TTS results are almost indistinguishable from natural speech -most of the time-.
   
           
      Since LFN is spoken as it is written it should be relatively easy to develop a text tot speech synthesizer for LFN.    
      However since spoken LFN is close to Italian, Spanish and Portuguese we can use the available text to speech software for these languages (all the software needed is available on the internet for free -legally-) .    
           
      The Italian and Spanish tts's produce nice results.    
           
         
           
      The results are nice but -according to experts- not perfect yet. To produce perfection we would need a custom made LFN tts.   www how to do this
           
     
Quick and dirty LFN
   
           
      We do not have to wait until a perfect LFN tts is available. There is a -quick and dirty- alternative: take one of the available TTS's (e.g. Italian or Spanish) and trick it into speaking better sounding LFN.    
           
      For example: a Italian TTS does not pronounce the X in the correct LFN way. But if we substitute all the X's by SCI, the tts produces acceptable X sounds. Likewise we substitute all CE with CHE , all CI with CHI, etc.
The end result is better sounding LFN. 		   
           
      Using different rules we can "mutilate" LFN text in such a way that Spanish tts produces acceptable sounding LFN.    
           
         
           
     
Make your own sounds
   
           
      Using TTS requires three kinds of programs:    
           
      1) A tts-shell program, below you find links to the two I use.    
     
Text To Speech Live Player. (Tip) Look beyond the somewhat wild interface and you will find nice functionality.
    
      ReadPlease 2003 plus a thirty day trial version.    
           
      2) A tts engine (for each language you need a different engine).    
      Free, Lernout & Hauspie Voices, (download here...). These tts engines can be played in both of the above mentioned tts-shells. I use the:    
      TTS engine with 2 L&H Italian voices(Male & Female) (Tip)    
      TTS engine with 2 L&H Spanish voices(Male & Female)    
      TTS engine with 2 L&H Portuguese voices(Male & Female)    
           
      3) Software for manipulating sound files    
     
To merge sound files or to convert to wav, mp3 etc and to add effects we need Sound Editing Software. There are many programs that can do this. I use: WavePad, Sound Editing Software and Switch Sound File Conversion Software, both good and free from NCH Swift Sound. (Tip)
    
           
      To make your own sounds, you should:
-download and install the: Text To Speech Live Player 
-download and install the L& H Voices: (download here...)
-Learn to use the Live Player software.		    
           
     

You can also use tts facilities found on the web.

Have al look at: the impressive AT&T Natural Voices™ Text-to-Speech Engine. You can use the free demo's of this program (demo1 or demo2 ).

Here is an other demo: provided by Acapela Group using High Quality TTS.
For demo and evaluation purpose only. Italian and Spanish.

An other one here (The background music you hear in the demo is not heard with the purchased Loquendo TTS. It is a copyright device to prevent illegal use of our product) Lots of Italian and Spanish voices (but with background music)

Demo Cepstral Voices, Try out a sample of some of the voices that we currently have available. We are building new synthetic voices for Text-To-Speech (TTS). Two Spanish voices, change the rate and pitch, and add effects.

Nuance RealSpeak™ Text-to-Speech solves business problems around the globe every day, by speech-enabling all sorts of customer service transactions on the telephone, in the car, on the desktop, on PDA's and other mobile devices. Experience its remarkable speech quality today, and check for yourself—is it real, or is it RealSpeak?

    
           
     

Very strange

Welcome to Vox Machina, the voice of the machine. Vox Machina is a web application that can convert your text input into a spoken word MP3 file.

try out the : Pipe organ and the cello voices.