Example of use of http://k.ylo.ph/2016/04/04/loudnorm.html A way to get a video with a better relationship of loudness for speech, music and "Action-Thunder" Perhaps all ways known, perhaps not. Example of video with ac3 + h.264 ffmpeg filter, /dual-pass/ mode : 1. Measurement  with results : ffmpeg -i "Desperado.ts" -c:v copy -filter:a loudnorm=I=-16:TP=-1.5:LRA=11:print_format=summary -f null - ... Input Integrated:    -26.0 LUFS Input True Peak:      -3.1 dBTP Input LRA:            18.1 LU Input Threshold:     -37.4 LUFS Output Integrated:   -17.0 LUFS Output True Peak:     -1.5 dBTP Output LRA:           11.8 LU Output Threshold:    -27.7 LUFS Normalization Type:   Dynamic Target Offset:        +1.0 LU "Desperado.ts" renamed to "Desperado .ts, 2. use of measurements  : ffmpeg -i "Desperado .ts" -c:v copy -af loudnorm=I=-16:TP=-1.5:LRA=11:measured_I=-26.0:measured_TP=-3.1:measured_LRA=18.1:measured_thresh=-37.4:offset=+1.0:linear=true:print_format=summary -c:a ac3 -b:a 448k -ar 48000 "Desperado.ts" Worth to try. regards