VSYNC OFF (without fps cap) vs VSYNC OFF (with fps cap)

[marci_g9x]

Hi!

I know game engines work differently so let's take just one game: Battlefield 4.

When i play with Vsync off and with uncapped fps (gametime.maxvariablefps 0) i feel heavy mouse lag. I know that the game has a 10hz tick rate which cause lag but, when i cap my fps below the average, it's super responsive (even at 30 fps). If i enable Vsync with capped fps, the responsiveness of the mouse is still better than with Vsync off without fps cap. So my question is, why is this happening in games and do AMD cards have lower lag?

I always preferred mouse responsiveness than motion clarity, so G-Sync is useless for me unless, they can remove these lags in Vsync off.

[nimbulan]

What are your system specs? The only reason I can see why you would be getting mouse lag with vsync off is that the game is using 100% of all your CPU cores and starving other processes (such as your mouse software.) How does it behave with vsync on?

[Chief Blur Buster]

The only reason I can see why you would be getting mouse lag with vsync off is that the game is using 100% of all your CPU cores and starving other processes (such as your mouse software.)

That's one cause.

There could be other causes, such as starving or slows down the game tick thread which then runs more lagged, resulting in less fresh input reads (mouse input, keyboard input, etc).

Other game developers who read these forums, can chime in!

[RealNC]

Note that the tickrate of 10 in BF doesn't mean you get input lag. That's just how many times the game state is synced with the server. Input happens at whatever FPS you get.

[Chief Blur Buster]

Note that the tickrate of 10 in BF doesn't mean you get input lag. That's just how many times the game state is synced with the server. Input happens at whatever FPS you get.

True, the game can be programmed that way. But who says all game programmers, develop it that way? -- I have a strong suspicion that sometimes gunshot trigger behavior appears tied into the tickrate in certain games. Probably not Battlefield 4, since a 10Hz tickrate would imply frequent >100ms input lag situations, which fortunately does not happen.

As an extreme example -- look at Skyrim for example. Its physics is so hopelessly tied to the frame rate; that glitches occurs when you try to run at frame rates above 60fps. So, obviously, not all games are programmed as framerate-flexible as they should be...

I should do some high speed camera input lag tests with different forced tickrate settings in some games that have customizable tick rates, and what latency changes occurs in which games.

[RealNC]

Counter-Strike 1.6 might be interesting to test this out. It supports tickrates up to 1000Hz (if your computer and OS setup can provide that; on Windows you used to need a utility running in the background that would put Windows into 1000Hz interrupt timer mode.) Newer CS games only go up to 128Hz, AFAIK.

[marci_g9x]

When the GPU uses 99% i feel the lag. So, when i cap the fps the gpu usage drop and everything is fine. I think GPU rendering is holding back because it's slow. Btw specs: 2500K, GTX 560 Ti

[nimbulan]

When the GPU uses 99% i feel the lag. So, when i cap the fps the gpu usage drop and everything is fine. I think GPU rendering is holding back because it's slow. Btw specs: 2500K, GTX 560 Ti

You might want to check your setting in the nVidia control panel for Maximum Pre-rendered Frames. Turn that down to 1 and see if it helps.