Your game runs at 120 FPS but still feels slow: how input lag works

Two different measurements

The counter says 120 FPS, yet the character turns a fraction after your thumb does. Motion looks smooth while the controls feel as if somebody wrapped them in a blanket. That feeling is input lag, and frame rate alone cannot tell you where it came from.

FPS measures throughput: how many complete pictures the game produces every second. Input latency measures duration: how long an action takes to travel from a button or mouse to a visible result. One number counts images. The other counts waiting time.

Higher frame rates usually shorten part of that wait because the game has more chances to sample an action and show it. They do not guarantee a responsive system. Frames can pile up before rendering, the television may process every image and the game itself may read controls late. A clean 60 FPS setup can therefore feel better than a badly configured 120 FPS one.

From button to pixel

Microsoft's Dynamic Latency Input documentation describes input latency as the delay between a physical press and the game reading it. Players care about the longer journey. The controller reports an action, the CPU updates the game, the GPU renders a frame and the display processes that frame before its pixels change. Every stage takes time.

Picture a supermarket queue. Your input may arrive just before the next frame begins or miss the cut and wait for another. If the CPU prepares work faster than the GPU can draw it, rendered work can form a second queue. NVIDIA Reflex paces those two parts so frames are prepared just in time rather than sitting around with increasingly old input.

The display is the final counter. Televisions often add motion smoothing, noise reduction and picture processing designed for films. Game mode removes much of that detour. Pixel response time is only the time a pixel needs to change colour; it is not a measurement of the entire button-to-screen chain.

More FPS helps, but it is not the whole answer

One frame lasts about 33.3 milliseconds at 30 FPS, 16.7ms at 60, 8.3ms at 120 and 4.2ms at 240. Raising frame rate reduces the interval in which a new action may be waiting, improves motion continuity and commonly lowers latency too. That correlation is real.

It becomes more interesting when the two variables are separated. In a SIGGRAPH Asia study led by NVIDIA researcher Josef Spjut, eight experienced esports players completed aiming tasks at different refresh rates with controlled extra delay. Removing 30ms clearly improved completion time, while refresh rates above 60Hz produced a smaller effect once their usual latency advantage was removed. It was a small aiming study rather than a universal ranking of displays, but it shows why smoothness and responsiveness deserve separate names.

Frame generation makes the distinction visible. Software-created intermediate frames can raise the displayed counter and make motion look smoother, but they do not contain a newer game input. Our DLSS, FSR and PSSR explainer covers that trade without mistaking generated pictures for native engine performance.

Ping takes a different route

Ping is the round trip between your machine and a server. Local input lag still exists with the router unplugged and a single-player game running offline. Online play combines both delays, which is why people often call every slow response lag even when the causes require different fixes.

Try an offline menu or camera. If that also feels heavy, inspect the device, rendering path and display. If the problem only appears as teleporting opponents, rejected hits or rubber-banding online, investigate the connection, packet loss and server. A faster monitor cannot move a server closer. A router restart cannot disable television processing.

How to reduce input lag

Start with the free changes. Enable Game mode on a TV and turn off motion smoothing, then make sure PC or console is actually using the display's highest supported refresh rate. Our guide to 120Hz and VRR on PS5 walks through output information, HDMI ports and cables. VRR matches display refresh to delivered frames; it does not manufacture the missing ones.

Use Reflex or Radeon Anti-Lag when the game offers a built-in option. AMD says Anti-Lag dynamically adjusts frame timing and works best when the GPU is the bottleneck. Without one of those modes, reducing ray tracing, shadows or resolution can give the GPU enough headroom to stop work piling up. Change one thing at a time instead of importing a driver's entire list of tricks.

Do not assume every wired controller must beat every wireless one either. Reporting rate, USB behaviour and the specific device all matter. Update its firmware, test both connections and compare the same scene. A magnetic keyboard can shorten key reset distance, as our Hall Effect keyboard explainer shows, but it cannot remove delay introduced by the display.

Test in this order: Game mode, correct refresh rate, the title's own low-latency setting, GPU load and finally the peripheral. A slow-motion phone recording that keeps the button and display in one shot is not a laboratory instrument, but it can compare two configurations. That is enough to find which change makes your setup respond first.

When the green counter says 120 FPS and the controls still feel heavy, stop chasing only the counter. Follow the action from button to pixel. The delay is waiting somewhere along that route.