Considering the great leaps and bounds made in resolution upscaling, why is framerate upscaling not getting the same love and progress? Why are we relying on TVs to scale down to 40 fps VRR instead of motion interpolating all games up to 60+?
Some thoughts:
TV motion interpolation has an advantage: Both frames already have motion blur in them, increasing smoothness and limiting chance of visible artifacts.
For games we have the advantage of motion vectors, but it does not help for things like transparency and moving shadows.
Upscaling (approx) 40 fps to 60 means an irregular distribution of generated and real frames. We get irregular sequences of r,g,r,r,g,r,g,r,r... Chances are this looks more smooth, but still stuttery due to this inconsistency, which would be better with MB.
The cost of interpolation subtracts from the frame times, so missing the target will happen more often.
Now compare the promise with movies at 24fps, which look smoother than games because they have MB.
So alternatively we could consider this as the target: 30 fps games with better MB.
To get this, we would store the previous framebuffers including motion vectors, which allows to resolve missing disocclusion data using the previous frame. (has this been done already? idk)
The cost would be constant and not more than 2 times of the current approximate solutions, but less than fps upscaling.
I assume this would feel better than messing with 40 fps, because the fixed 1:2 ratio.
Shadows and transparency is still a problem.
Would be interesting to see what's the better compromise. May depend on the game.
Personally i think we just need those 60 fps for games. Upscaling only makes sense to achieve rates above that, to drive high Hz displays.
Would be nice to see some 30->60 fps videos using DLSS3, but then i would be still unsure if it only looks smoother, or also plays better as well.