1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

Why capping frame rates induces latency/input lag?

Discussion in 'Videocards - AMD Radeon Catalyst Drivers Section' started by Eastcoasthandle, May 21, 2018.

  1. Eastcoasthandle

    Eastcoasthandle Ancient Guru

    Messages:
    1,893
    Likes Received:
    88
    GPU:
    R9 Fury
    I've noticed that even though that if you cap frame rates to either refresh rate or some random number induces input lag. Using MSI A/B it shows up as very small squiggly line. The game is still smooth but you can tell something is off.

    I recall that part of capping framerate popularity was do to Main Menu screens where you can have heat runaway as frame rates would be 350-500 FPS depending on game etc. But if you cap frame rates at 300 fps for example I've noticed input lag.

    But for online games now a days I really don't see how one gets an advantage in BF1, Fortnite, COD WW2, Overwatch CS: GO, etc. In fact you might be at a disadvantage. As there is an odd Out of Sync behavior I notice with mouse/keyboard.

    So can someone tell me if they tried capped vs uncapped what their experience has been?
     
    Last edited: May 21, 2018
  2. slash_One

    slash_One Member

    Messages:
    31
    Likes Received:
    13
    GPU:
    ATI HD6770x2
    I fixed input lag for some games with d3doverrider on win7 x64. Point was to enable tripple buffering with vsync on, and ingame framerate cap = screen refreshrate...Not always works:cool:
     
  3. RealNC

    RealNC Ancient Guru

    Messages:
    2,499
    Likes Received:
    737
    GPU:
    EVGA GTX 980 Ti FTW
    If the game maxes out the GPU when running uncapped, frame limiting will actually reduce input lag. For games like CS:GO that are very light on the GPU and thus get CPU bottlenecked this doesn't apply. But for GPU heavy games, frame capping helps with input lag.

    Also note that an input lag reduction can result in a feeling that something's off. You can get less input lag but feel like you have more, because you're used to slightly higher lag. Unless you do a measurement, you can't be sure.

    Again though: this is true for GPU-bound games. CPU-bound games behave as expected: the lower your FPS, the higher your average input lag will be.

    And of course, it also depends on the difference between capped and uncapped. If uncapped a game runs at 400FPS and you cap it to 60FPS, then yeah, don't expect the same input lag... 400FPS means 2.5ms frame time, and 60FPS 16.7ms frame time. Even if frame capping would provide an input lag reduction, it's offset again by the frame time difference.
     
    Eastcoasthandle and Noisiv like this.
  4. Eastcoasthandle

    Eastcoasthandle Ancient Guru

    Messages:
    1,893
    Likes Received:
    88
    GPU:
    R9 Fury
    ^
    Good info thanks.

    I keep reading to only frame rate cap using the game whenever possible. That AMD, MS AB, etc induce latencies while games that have the ability to limit frame rates don't.
    Is this true?
     

  5. RealNC

    RealNC Ancient Guru

    Messages:
    2,499
    Likes Received:
    737
    GPU:
    EVGA GTX 980 Ti FTW
    In-game limiters have the potential to reduce input lag. But they can have micro-stutter.

    External limiters like RTSS will not reduce input lag. They will instead give you the natural latency of the target frame rate (100FPS will give you 10ms for example.) The exception is when your GPU is maxed out when leaving it uncapped. In that case, an RTSS cap will reduce input lag because it prevents the GPU from being fully saturated. RTSS is very accurate and thus can get rid of micro-stutter. It is even accurate enough to use with vsync in order to get "low latency vsync" (see this and this.)

    I don't know how AMD's limiter behaves. NVidia's will add some input lag.
     
    OnnA and Eastcoasthandle like this.
  6. Eastcoasthandle

    Eastcoasthandle Ancient Guru

    Messages:
    1,893
    Likes Received:
    88
    GPU:
    R9 Fury
    Thanks RealNC I appreciate your knowledge on the subject.
    I tested it out on BF1 myself and can confirm that using MSI AB does not hinder gameplay with added latency nor input lag.
    I used 150 FPS as a target and I adjusted IQ settings accordingly to get it 150 FPS.
    Game play was smooth, my controls were timely and I seem to get the same information about what's in front of me in the game as the server.
     
  7. Pyrage

    Pyrage Master Guru

    Messages:
    372
    Likes Received:
    24
    GPU:
    580 CF@ 1470
    I use Freesync so sometimes I can't mess with RTSS because the game will feel kinda weird, it feels fast but it's not smooth, even though Afterburner shows me that the frametimes are even. It also adds weird micro tears depending of the game. So I just mess with it if I feel that I need to.
     
  8. RealNC

    RealNC Ancient Guru

    Messages:
    2,499
    Likes Received:
    737
    GPU:
    EVGA GTX 980 Ti FTW
    Have you tried enabling vsync? Freesync behaves very similar to g-sync, and there, the "perfect" setup is vsync ON and RTSS with an FPS cap of -3FPS below max refresh (in this case, -3 max freesync Hz.)
     
  9. Pyrage

    Pyrage Master Guru

    Messages:
    372
    Likes Received:
    24
    GPU:
    580 CF@ 1470
    That's how I do it, but it doesn't matter. Some games will work fine, others feel weird and the only solution is to either stop using RTSS and use the ingame FPS cap, or not using any caps at all and just leave it with Vsync on only.

    In WoW micro tearing is everywhere if I use RTSS, capping at 141, 142.143 etc nothing works. The game only becomes smooth and tear free if I use the ingame cap setting it to 141.
     
  10. Eastcoasthandle

    Eastcoasthandle Ancient Guru

    Messages:
    1,893
    Likes Received:
    88
    GPU:
    R9 Fury
    [​IMG]

    So yeah, MSI AB works pretty well.
     

Share This Page