Many people believe that if they will process their video, it will look better than the original.
There’s also this absurd term that was coined – “video enhancement”. You know how in the movies, the technician in front of the surveillance video, looking at a completely blurry license plate, is being asked to enhance it? Then the image magically sharpens so you can see clear digits. Well, that’s really cool but unfortunately impossible.
Remember how Deckard is zooming and panning on a photo in Blade Runner? that was shot in 1982 and It’s still science fiction 31 years later. Sorry.
The golden rule that all hardcore video guys know is simple. ANY digital video manipulation degrades video quality. Period. Yes, even if the result does look better to YOUR eyes. And that’s the keyword here – YOUR eyes. video viewing is entirely a subjective experience. What looks good to your eyes, will not necessarily look good to someone else.
There is no magical way to recreate pixels that were not captured during the filming or animation creation process. Sure, there are algorithms out there that do a really nice job of creating those needed pixels in processes like upscaling or deinterlacing, but this is done by interpolation. For example when you upscale an SD video to HD, it is done by creating a new pixel which has the average color information of its neighboring pixels. (a very simplistic and very inaccurate explanation, but that’s basically it).
So unless your video processing software is that clever that it can transport itself to the time and place where that shot has taken place, re-shoot it with a better camera, then go back to your computer with that new shot – then it’s degrading your video quality.
So my recommendation is – Avoid any unnecessary video processing. Focus on better video encoding instead. Learn how to encode your video more efficiently (or leave it to the pros).