i tried finding .wem openers on the internet, but all of them cant extract the sounds from the files, for some strange reason, wanted to post knight paladin voice onto youtube because of how epic it is
I converted everything.
I used this .bat that I placed in the sox folder
FOR /f "tokens=*" %%A IN ('dir /b .\dow3*.wem') DO sox --type raw --rate 22050 --encoding signed --endian little --bits 16 --channels 2 .\dow3\%%A "converted/%%~nA.wav"
/f means process the output of the command in single-quotes (inside the parens) one token at a time. "tokens=*" means parse an entire line of output as a single token. (Basically, it means whitespace is not considered a delimiter.)
'dir /b .\English(US)*.wem' means list all *.wem files in the dow3 directory without date, size or other metadata.
takes all the .wem in the '/sox/dow3' folder and turns them into a .wav in the '/sox/converted' folder
make sure you don't have parentheses in your folder name, was messing with my command. You can get SoX for free (http://sox.sourceforge.net/).
Looks like some files are 44kHz and other are 22kHz tho. So some sound really slow with the 22k sampling I used.
Problem now is that I still have 17300 files to sort by faction. Would be nice to know which file is what.
I made a .wem hex reader and I now know the true encodeing the audio files are supposed to have: 48000 mono and not 22050 stereo. There is a click at the start of the files tho. Comes from the way I use SoX to convert the .wem into .wav. I tell it to import a raw fille (only audio data) and add a self describing header (with size, bitrate, channels...). Thing is, this hear is already there, so we end up with two headers, the second one being interpreted as audio data (click!). Might have to make a program that removes the first header for all the .wem files, before running a SoX conversion with the proper params.