I already saw a few post regarding training acoustic Model with Simon.

I wanted to use Ralf Herzogs Dictonary with the current german voxforge Modells but of course a lot of triphones are missing. Since I working for a small film company I got a lot of speech samples on my hand and like to train my own or adapt the existing german acoustic modell.

I tried to use google.
But although some people mention they use simon for the training noone really explains how it works step by step for a small example. (Or maybe I'm just stupid and can't find it in google)

I found some very old descriptions but I'm not sure if all the steps are still necessary. One wants me to install KDE "emerge" for example.

Any pointers would be helpful.

Binh