[mary-users] HMMVoiceConfigure

[Thorsten Westermann]

Hello!

I am having a problem again.

This time in HMMVoiceConfigure.

The error description is:
------
Checking directories and files for running HTS training scripts...
Problem with wav and data/raw directories: wav files and raw files do not exist.
Problem with wav and data/raw directories: wav files and raw files do not exist.
Problem with transcription directories text or data/utts (Festival format): utts files and text files do not exist.
 the transcriptions in the directory text will be used to generate the phonelab directory, if there are no data/utts files(in Festival format), please provide the transcriptions of the files you are going to use for trainning.
Problems with directories wav, text or data/raw, they do not exist or they are empty.
------

This error may be related to the fact that I don't understand how to set up the variables for the HMMVoiceConfigure.

In the documentation it is said:
------
  HMMVoiceConfigure.dataSet      =  german_set_name
  HMMVoiceConfigure.featureList  =  feature_list_de.pl  (the set of context features used for this voice can be change in this file).
  HMMVoiceConfigure.speaker      =  speaker_name 
  HMMVoiceConfigure.lowerF0      =  40  (male=40,  female=80)  
  HMMVoiceConfigure.upperF0      =  280 (male=280, female=350)
  HMMVoiceConfigure.voiceLang    =  de
------

My questions would be:

1) What exactely should be used as the variable for "german_set_name" when I want to create a German voice?

2) The variable "featureList" does not show up in the current version beta version (I did not use the SVN but the compiled beta version).
Is it a problem that "featureList" does not show up in the settings window?

Thank you for the help again.

MfG
Westermann
-- 
Jetzt kostenlos herunterladen: Internet Explorer 8 und Mozilla Firefox 3.5 -
sicherer, schneller und einfacher! http://portal.gmx.net/de/go/atbrowser