Thanks for the clarification!
Let me explain briefly how do I use LumaFusion:
- I edit 100% of my videos using LumaFusion. I publish usually 1 video a week up to 30 videos a month. I tried desktop NLE's but I'd rather make the most of my commutes, than waiting to edit back home.
- Even though I love LumaFusion, I don't love editing. At least, when it comes to repeating the same moves again and again...
- I plan to produce a series of courses with the same setup: a front facing camera recording my voice, an overhead camera pointing to a whiteboard and my computer screen being recorded.
What do I want to accomplish? I would like 90% of the work to be done programatically, without me having to do it by hand.
How I would like to do it? This is my plan:
- I would AirDrop/copy all media into my computer. I would create a new folder and put videos inside. I would rename them front.mov, overhead.mov and screen.mov. I will also copy backgroundmusic1.wav, etc.
- I would launch a script in my computer that would put front.mov as the track number 1. It would also set audio to principal and then apply favorite LUT.
- Script then will capture the first frame of the overhead camera and use OpenCV to detect largest blob (whiteboard surface) I would detect changes on the blob surface to detect when I'm writing, using background subtraction. Every time it detects I'm writing, it would extract that fragment into a clip, taking the previous 1,5 seconds. Each of these clips would go to track number 2.
- Script then will check for movement on my captured screen and do the same thing. Extract moments were I'm moving around the script and extract those. These clips would to track number 3.
- Video tracks 2 and 3 would be muted.
- Background music would be copied to the audio track 1 and configured for ducking.
- When I say it would go there or that, I mean: these changes will be saved on the archive.spryarchive, according to the specs.
- I would write the time codes for in/out for each clip, project parameters, etc.
- I would put this archive.spryarchive in the same folder as the files, zip it, rename as foldername.spryzip and Airdrop it to my iPad for final edits.
So that's my goal! I've been checking several python libraries to manipulate the videos and it seems doable. This workflow would be highly beneficial, given it would save me around 2-3 hours per video (For a 30 days, 30 videos publishing schedule, that's taking back 90 hours back of your month).
Biggest challenge right now, how to add this information to a well structured spryArchive file