just so i have this next time i want to dub some audio with ffmpeg

ffmpeg.sh

# mix two audio streams together:
youtube-dl "https://www.youtube.com/watch?v=c38HJR-9vhU" -f best -o burnafterreading.mp4
youtube-dl "https://www.youtube.com/watch?v=Pqh63ca__mM" -f best -o chromatica.mp4
ffmpeg \
  -ss 77.2 \ # skip first 77.2 seconds of first input
  -i burnafterreading.mp4 \
  -i chromatica.mp4 \
  -filter_complex "[0:1][1:1] amix=inputs=2:weights=2 1" \ # mix track 1 (audio) of inputs 0 and 1, make audio 0 twice as loud as 1
  -map 0:0 \ # use track 0 (video) from input 0
  -c:a aac -strict -2 \ # don't think this is necessary, default encoder settings should be fine
  -t 49 \ # trim output to 49 seconds long
  output.mp4
  
# replace audio:
ffmpeg -i input.mp4 -i input.wav -c:v copy -map 0:v:0 -map 1:a:0 -c:a aac -shortest output.mp4

# crop and overlay a transparent png:
ffmpeg -i fergorig.mp4 -i wombo.png \
    -filter_complex '[0:v]crop=240:240:80:0[a]; [a][1:v] overlay=140:0' \
    -pix_fmt yuv420p -c:a copy out.mp4