Split audio/video streams from single source

This is a quick hack to split a single video file (e.g. of a mixed interview/observation session) into separate audio-only and video-only files based on time codes.

Usage

Given a source audio video file named input.mov and a text file containing time codes and section names named times.txt:

Interview	0:00	4:21
Sibling Harmony	7:06	12:58 complete
Fly Away Blimp	13:47	18:44 complete
Get Tim Home	18:58	21:06 complete
Interview	21:07	24:40
Cannonball!!	24:48	37:36 complete
Excavation Site	37:55	1:02:06	complete
Interview  	1:02:07	1:06:50
Final Countdown	1:07:01	1:32:29	incomplete
Interview	1:32:30	1:42:05

run

$ python split.py input.mov < times.txt

The output directory data/output must exist.

Segments with the name 'Interview' will be split into audio-only files while other segments will be split into video-only files.

Input File

The general format for each line of the input file is

Section Name    start_time  stop_time    meta_data

start_time must be in either HH:MM:SS or MM:SS time format. stop_time must be either in one of those time formats OR it may be - which is taken to mean 'end of the input file.' The meta_data field is currently ignored but could be used for simple quantitative analysis based on keywords entered for each segment.

Todo

• Use argparse to create a more flexible command interface
• accept command line option to select output directory
• accept command line options to transcode the video/audio into codecs different from the input codec.