No No...I'm always keeping the original vintage video...just "restoring" like before it was mosaiced and then applying the sub and processing it. Sorry if I didn't make that clear. So that when LADA II is released I can do it all over again! SIGH!I meant this: https://en.wikipedia.org/wiki/Generation_loss
Say you have a video. You run your AI de-mosaic magic on it. Aside from what changes you intentionally do to the video (de-mosaic) it will involve a re-encode with a lossy encoder like x264 or hevc or what have you. Every time you do that, you lose some of the original information in the video.
So what I meant was, if at some point later you want to apply your newer, better demosaic algorithm, you may wanna do it on your originals, and not (again) on the ones that had already been processed.
This actually may be of no concern to you, but it's a thought that crossed my mind upon reading your plans towards your library
P.S. thanks for the subtitle work you (and others here) do, my hat's off to you.
Last edited:
