v1.5.7

A few minor changes:

• Added --re-encode-audio boolean option, see #31
• Added --threads <int>, --threads-input <int> and --threads-encoding <int> options to pass through to ffmpeg's -threads arguments (see #30); --threads-input and --threads-encoding take priority, but --threads will be used for either one of the others that isn't specified
• minor tweak to swears.txt

v1.5.6

fix issue with getting channels

v1.5.5

Pull from :6.0 rather than :latest image for ffmpeg binaries.

v1.5.4

Make padding a floating-point value (e.g., 1.5 seconds).

v1.5.3

restore pairwise for older python 3 versions

v1.5.2

Removed a leftover volume filter that should have been removed when I added the afade stuff for #13.

v1.5.1

Default pad to "0" and set both fade in and out times to 10ms

v1.5.0

Attempt at fixing #13, using afade filter instead of volume on/off filter

v1.4.0

Several improvements:

• fix bug where the last subtitle entry would be lost
• ignore -d/--downmix downmix if the audio track(s) are already just stereo
• fix extraction of srt file embedded in video (and allow specifying index e.g., with -l eng:3 vs just -l eng)
• reduce verbosity of output of internally-executed ffmpeg commands

v1.3.1

Added --json option to write a JSON file with information about input, output and edits.