This will aim to answer questions about Street Fighter 6 input lag testing. You may be familiar with my regular uploads to https://twitter.com/noodalls or may have seen my contribution to Digital Foundry here https://youtu.be/U3WrOySoiRk?si=3zoFN6raptGpsYer

First let's look at Street Fighter 6 on the PS5. The video is captured using a Sony ZV1. This allows for 1000fps, but only for 3.33 seconds. I have set up the Arduino to press the right direction for 49ms then release for 50ms. The LED will behave in the same way. This should mean for 3.33 seconds we get approximately 33 tests.

Controller was Brook UFB fusion. IDR was on. Right direction was used as it provides a reliable transition between white to black which is easy to pick.

These settings will be consistent across the various different tests. 

Below you can see the results. Results averaging 56.8ms, which is fairly much in line with prior results. Now, you may notice that we are using the second command on the command history. We will get back to why later. 

This second video is the same setup, but this time recording at the bottom of the screen, or more precisely the second bottom command. The results for this are 61.6 ms. 

Now, why did we choose the commands the way we did? Why not the top and bottom command history icons. The reason for this is the second from the top icon is 25% of the way down the screen (274/1080 = 25.3%) and the second from bottom icon is 75% of the way down the screen (818/1080 = 75.7%). Now, given it's 56.8ms a quarter of the way down the screen, and 61.6ms, it would follow that it takes 4.8ms to travel half the screen distance. This would fit with it taking 8.3ms (near enough to +-1ms) to travel the whole screen.

So, more precisely, you could call PS5 SF6s input lag 54.7ms at the top of the screen, or 63.6ms at the bottom of the screen. For comparisons between versions of SF6 this is not as important, so long as the area used is consistent. However for comparisons between different games, ideally it should be fair.