A look at using Ultimate Voice Remover, a free frontend for AI audio source separation models, to remove background music from TV clips and radio broadcasts. Then, using FFMpeg to separate audio tracks, as well as recombining single and multiple audio tracks back into a video using FFMpeg.
Ultimate Vocal Remover GUI GitHub:
https://github.com/Anjok07/ultimatevocalremovergui
FFMpeg Windows Builds:
https://www.gyan.dev/ffmpeg/builds/
FFMpeg Commands:
extract audio:ffmpeg -i input.mp4 -vn -acodec pcm_s16le -ar 44100 -ac 2 output.wav
combine 1 audio track:ffmpeg -i "video.mp4" -i "audio.wav" -c:v copy -c:a aac "output.mp4"
combine 2 audio tracks:ffmpeg -i "input_video.avi" -i "track_1.wav" -i "track_2.wav" -map 0:v -map 1:a -map 2:a -c:v copy "output_file.mp4"
Subreddit Surfing with Cardiff Electric and Vinnie Paulino
https://www.youtube.com/@subredditsurfing