A number of users report that OmegaT does not properly recognize text file encodings.
The problem is that they interpret the [auto] of the text file filter as meaning [automatic recognition].
It would be less troublesome to have [os default] as the encoding value when what OmegaT does is use the OS default encoding, and eventually [declaration] when OmegaT takes it data from the file itself (encoding declaration etc).
Text filter/PO filter/Key-value filter/HTML-XHTML/Subrip/Latex : change [auto] to [os default]
all the other [auto] (incl.XML formats) : change [auto] to [declaration]
This will go a long way easing the processing of text based formats in OmegaT.