Feature request: expose splitter parameters to args

[xenotropic]

I initially just tried the instructions (put long files in put_audio_files_here), and got nothing, just an output line of omitting [filename] for each of my two files. Upon investigation, it seemed like I needed to make the splitter parameters more sensitive: to split on less-quiet and/or shorter silences. To do that I have to edit those params in two different places. It would make the tool more usable if those parameters (minimum silence length and threshold) were arguments that could be passed in to audiosplitter.py and/or pipeline.py.

[osok]

I have this addressed in my fork. I already made one pull request. Once that gets accepted I'll make another to add this functionality. Meanwhile you can grab the code from my repo.