delay-audio-with-ffmpeg.sh

# use this command to add 5-seconds delay to the audio track of a video
#
ffmpeg -i input.mp4 -itsoffset 5 -i input.mp4 -map 0:v -map 1:a -vcodec copy -acodec copy delayed.mp4

merge-video-files-with-ffmpeg.sh

# successfully tested on mac os' bash terminal
# if needed, you can download ffmpeg from http://www.ffmpeg.org/download.html
# 
ffmpeg -f concat -safe 0 -i <(for f in ./*.mp4; do echo "file '$PWD/$f'"; done) -c copy output.mp4

For reference, on video recordings of Chapelle des Lombards:

1. extract the audio with ffmpeg -i harissa.mp4 -vn -acodec copy harissa.audio.aac, or in pcm/wave format (for audacity): ffmpeg -i harissa.mp4 harissa.audio.wav,
2. sync & mix extracted audio with audio from secondary recorder, using audacity (cf Sync-Locked Track Groups - Time Stretching - Audacity Development Manual and Syncing two tracks: methods, features and suggestions - Audacity Forum)
3. re-mux the exported audio mix with the video, with an audio delay of 0.1 seconds: ffmpeg -i harissa.mp4 -itsoffset 0.1 -i harissa.mix.m4a -c:v copy -map 0:v:0 -map 1:a:0 harissa.mix.mp4
4. I split into one video per song using https://github.com/mifi/lossless-cut, or:

ffmpeg -i harissa.mix.mp4 \
-vcodec copy -acodec copy -ss 00:03:57.701 -to 00:10:13.932 harissa-01.mp4 \
-vcodec copy -acodec copy -ss 00:10:13.932 -to 00:15:19.000 harissa-02.mp4 \
-vcodec copy -acodec copy -ss 00:15:19.000 -to 00:19:44.697 harissa-03.mp4 \
-vcodec copy -acodec copy -ss 00:19:44.697 -to 00:28:54.213 harissa-04.mp4 \
-vcodec copy -acodec copy -ss 00:28:54.213 -to 00:35:08.469 harissa-05.mp4 \
-vcodec copy -acodec copy -ss 00:35:08.469 -to 00:41:45.105 harissa-06.mp4 \
-vcodec copy -acodec copy -ss 00:41:45.105 -to 00:52:06.713 harissa-07.mp4 \
-vcodec copy -acodec copy -ss 00:52:06.713 -to 00:59:31.303 harissa-08.mp4 \
-vcodec copy -acodec copy -ss 00:59:31.303 -to 01:07:07.241 harissa-09.mp4 \
-vcodec copy -acodec copy -ss 01:07:07.241 -to 01:15:44.531 harissa-10.mp4