Native 4K or Checkerboard "uprendered" 4k

[Guitarguy]

SvennoJ said:

Chevinator123 said:

Im still a bit confused about the checkerboard thing. What exactly is the pro rendering? Specifically what does "2 1080p frames" equal? What is the equaled res that its upscailing from to 4k?

IM SO DAMN CONFUSED

Imagine a checkerboard, black and white squares alternating.

Frame 1, you render the black squares
Frame 2, you render the white squares

To make a full picture, take the rendered squares, interpolate the non rendered squares using the rendered squares and info from the previous frame when possible.

So basically you render 1920x2160 or 3840x1080 depending on how you look at it. Since the squares are offset each line it's better than interlaced rendering yet still exactly half the number of pixels of 3840x2160.

I don't think this is how it works. What you are referring to is similar to field rendering which is what the early PS2 games used(also why alot of early PS2 games looked jagged). It renders two alternating fields of scanlines(odd and even) per frame and combining the two to create a smooth motion. This allowed developers to render games internally at half the resolution(say 240i/p) and output at 480i/p(the standard back then) with enhanced frame rates.

https://www.youtube.com/watch?v=TGTR0G2xC1E

Checkerboard rendering is this:



It essentially  averages the neighboring two pixels(diagonal from eachother) and reconstructs one parallel to both. I think this is the basis of it, although I have read it reconstructs 2 pixels for every 2 native pixels which makes sense as 4K contains 4 times the amount of pixels as 1080P(1080P is 1920 by 1080 whilst 4K is 3840 by 2160). There are no alternating pixels popping in or out like with field rendering. So you fill in the black empty areas with reconstructed pixels, giving you 4 times(as 4K/2160P has 4 times the amount of pixels as 1080P) the amount of pixels at much less expense to the GPU compared to native 4K