What are the best practices for scaling down framerate of a static video. For example, consider some footage captured at 360 FPS from a videogame. How would one scale this down for a 60 Hz display or a 30 FPS YouTube video? Is it possible to mix several frames into one in a way that a person watching the video would find satisfying or does throwing away the extra frames give the best possible result?
There are only two options to see every frame; play it slower, or get a display that can show it at its native frame rate. The only other way, is to ignore the excess frames.
And the second option only works if all of the components in the signal chain can handle such a high frame rate. For example, if YouTube doesn’t support 360 FPS, you’ll only be able to see the max rate they support. In fact, they’ll probably downconvert it to preserve bandwidth. You don’t have any control over what processing YT applies to your uploads.
For the first method, in most NLE’s, you would change the “Clip Attributes” (Resolve nomenclature) from interpreting the clip at 360 FPS, to your system’s maximum output rate. If that’s limited by a 60hz monitor refresh, then 60FPS.
This will slow down the clip enough for your system to show every frame. The amount of slowdown depends on the rate of original capture vs your systems max rate. For example capturing at 360, but displaying at 60 is 60/300, or about 17% of the original speed.
So another easy way to balance the equation closer to real-time is to limit your capture rate to the same as your display rate. In other words, if all you can display is 60FPS, then capture at 60FPS if you want 1:1 playback.
There is one scenario where capturing more frames than you need is practical, but it’s used more in VFX pipelines than video game streaming. Capturing high frame rates with high-end cameras yields less motion blur than typical frame rates. This makes tracking, keying, and rotoscoping for VFX more accurate, and the results can be blended together to re-create natural(ish) motion blur.