Quote:
Originally Posted by pip08456
To be honest there is no software that can do that. The only thing any software can do is make a "best guess" at what the missing pixels are.
|
True. There is an old saying in the Computer Industry that applies especially to digital video. Garbage in, Garbage out. Digital Video compression systems can work in two ways. Lossy, and Lossless. Lossless video compression systems still produce huge files, so their use is generally limited to professional production houses.
99% of consumer video is produced using Lossy compression. This reduces the size of the video considerably, but it loses a lot of data, hence the name. People also often reduce the resolution of the video to reduce the size. Again, the computer does this by losing pixels.
Unfortunately, as noted by pip, there are various products that can scale the video up, but these work by inserting extra pixels that are essentially a best guess at what was there. They are not usually that accurate.
This is why production houses and media companies prefer to do all their editing on uncompressed video. This is rather large. If they are using the standard 8 bit colour depth (8 bit red, 8 bit green, 8 bit blue, making 24 bits) which is the standard for monitors, the data will be around 9.3 GB per minute of video. If they switch to a 16 bit colour depth (which is standard for films, and presumably HDR video), the data is 18.7GB per minute. Yes, that is Gigabytes, not bits. For a 2 hour film, you'd be talking around 5TB of video.
More often than not, they use a lossless compression system, such as Apple's ProRes, which loses almost no quality, but brings the data required for the above mention 8 bit per channel video down to around 1.7 GB per minute.
They do not generally edit videos with lossy compression, purely because when you compress with a lossy codec, you always lose some quality. That's why it's best to start with the best quality possible.