Hi Ralph,

I'd love to spend more time submitting my own speech. 

But if you want to get technical about it ... ... I manually processed all users who submitted audio books - depending on the audio book this about 1-2 hours of work each (mostly adding new words to the pronunciation dictionary - it's currently a semi-automated process ... working towards being fully automated). 

Writing scripts, replying to forums posts and just keeping the website going actually takes up a fair bit of my time.

Right now there is no limit on the amount of speech a specific person can submit.  Submissions with lots of different words will help improve our triphone coverage.  If you get bored submitting your own speech, you might be interested in processing audio book speech - let me know if you are interested.

Ken 

Hello Ken!

OK, you do all the hard work.  I understand that.

I will try to submit speech with a good coverage of different words.  This is my concept when I submit in the German language.  I try to cover different topics so it should be possible to get a not too bad language/grammar/speech model.  

I am not getting bored submitting my own speech.  I'm just getting started. I would like to get involved more in the project, but it takes time.  I have installed Cygwin, HTK and Julius, but until now I haven't been successful.  I will try again and again until it works.

So I am trying to learn something about your project.  I have started reading the HTK book.

You know, I want the German section to get worked.  And your tutorials are quite good.

Greetings, Ralf

Update: Max submissions from one particular person should be no more than 2-3 hours, otherwise the acoustic model gets biased to that person's voice.