I tend to think of it similar to supersampling. (In short its a technique used where you render the video in higher scales, and then scale it down in native resolution).
Streaming services if they have captured 4k they compress it and downscale it to all other resolutions so they dont have to send much data. But the GPU probably does the job better when it has access to the raw video.
Also you might hear “compression”, but in reality data is thrown away in a non recoverable way
Latest Answers