I wanted to save the “workflow” I used for my last video.
Video? Oh yeah, I started a new Youtube channel. It’s just for practice. I want to get a cam-audio guitar practice vid up as close to daily as possible.
The channel is here: https://www.youtube.com/channel/UCmthqO1oB5PKlfZSwbKn_0Q
I also have an old channel that I still upload to, in case you were wondering: https://www.youtube.com/channel/UC-qXuDwTbmeJFJb0WZvM4Kw
The first video I put up I used my full-featured nonlinear video-editing software of choice, cinelerra. But all I even *used* it for was to crop the beginning and end!
That first video is here: https://www.youtube.com/watch?v=AO4GeXrR8_0
I’m practicing an acoustic version of “The Suburbs” by Arcade Fire. I feel like the world is going to come to an end in a few days or weeks or months or years and in that mindset I am beginning to reinterpret some of the lyrics to this song in new ways.
For the second video, I was smarter, and did everything in the command line.
The second video is here: https://www.youtube.com/watch?v=6T-fTcca6cU
I’m practicing “Suzanne” by Leonard Cohen. I’m pretty torn up about his passing and I want to learn his songs because I respect him so much.
Walker’s instructions for how to make simple videos:
- Record the video on your Flip cam.
- Get the file into a linux or mac computer with ffmpeg installed.
- Use the following command to take the audio out of the video:
ffmpeg -i videofile.MP4 audiofile.wav - Open the audio file in audacity. Apply noise reduction, then very light compression, then very light equalization (roll off extreme lows). Export to newaudio.wav
- Use the following command to remux the old video with the new audio AND trim the video:
ffmpeg -i videofile.MP4 -i newaudio.wav -map 0:v -map 1:a -codec copy -ss 00:00:00 -t 00:04:00 remuxedvid.mkv
Take note that one should replace “00:00:00” with the intended start time. Everything before that will be trimmed. “00:04:00” should be replaced with the total time. Everything after [start + total] time will be trimmed as well.
Compared to the first time around, I avoided 15 minutes of rendering and 15 minutes of reencoding. So using ffmpeg as much as possible and cinelerra as little as possible is usually the most time-optimal solution.
Also, by being sneaky I did all of this using the flip cam’s space. No need to delete temp files from two locations.
Walker’s new rule for doing video work: Do you need to composit something? Then use cinelerra. If you do not, then do not use cinelerra.
