What is input lag and why does it matter for FPS games?
Input lag — also called click-to-photon latency or end-to-end latency — is the total time between physically pressing your mouse button and the corresponding pixels changing on screen. Every link in the chain adds milliseconds: switch debounce (1–4 ms), USB polling wait (1–8 ms), game engine processing (1 frame at your FPS), GPU render (1 frame), frame buffer queue (0–2 frames), monitor scanout (half the refresh period on average), and panel response (1–10 ms GtG). A competitive setup targets under 25 ms total; consoles often sit at 60–100 ms.
How does monitor refresh rate affect input lag?
Each new refresh is your only opportunity to display an updated frame. On a 60 Hz panel the next frame is at most 16.7 ms away and on average 8.3 ms away; on 240 Hz those numbers drop to 4.2 ms and 2.1 ms; on 360 Hz to 2.8 ms and 1.4 ms. Higher refresh also shortens the scanout window — top-of-screen pixels light up earlier relative to the click. The total saving from 60 Hz to 240 Hz is roughly 12–15 ms of average input lag. Verify your real refresh at /monitor/refresh — sometimes Windows defaults to 60 Hz on a 240 Hz panel.
Does mouse polling rate really matter?
Yes, but the diminishing returns are steep. At 125 Hz polling the worst-case wait for the mouse to report a click is 8 ms (average 4 ms). At 1000 Hz it drops to 1 ms / 0.5 ms. At 4000 Hz (modern wireless flagships) it drops to 0.25 ms / 0.125 ms. Going from 125 Hz to 1000 Hz saves a meaningful 3–4 ms; going from 1000 Hz to 4000 Hz saves under 0.5 ms and stresses your CPU. Confirm your actual polling rate with /mouse/polling before paying for a 4K-poll mouse.
What about wireless vs wired latency?
Modern 2.4 GHz gaming wireless (Razer HyperPolling, Logitech Lightspeed, Pulsar) is statistically indistinguishable from wired — sub-1 ms median with very small variance. Bluetooth is a different story: HID over Bluetooth aggregates reports into 7.5 ms intervals, adding 4–10 ms of variable lag, and is unsuitable for competitive play. Wireless 2.4 GHz with the receiver close to the mouse (USB extension to the desk) avoids any radio retransmits. Battery low-power modes can throttle polling — check /mouse/polling while gaming, not idle.
How much does my FPS affect input lag?
A lot. At 60 FPS a single frame takes 16.7 ms, so the engine adds 16.7–33 ms (one or two frame queues). At 240 FPS each frame is 4.2 ms. At 500 FPS it's 2 ms. Competitive shooters like Valorant and CS2 cap well above 300 FPS on modern hardware, and you should run uncapped if your monitor supports it. Reflex / Anti-Lag features cut the render queue from 1–2 frames to ~0.5, saving another 5–15 ms. The single best latency upgrade for most players is raising FPS, not new hardware.
What is click-to-photon latency and how is it measured?
Click-to-photon is the gold-standard input lag metric — the wall-clock time between mouse switch closure and the first pixel changing on screen. It's measured with a photodiode taped to the monitor and a logic analyser on the mouse switch, captured at microsecond precision. Reviewers like NVIDIA LDAT, Rtings, and Hardware Unboxed publish click-to-photon for popular monitor + mouse combos. Software estimates (including this tool) sum the known additive bounds of each chain link; they're typically within 5 ms of the photodiode result for well-characterised hardware.
Should I turn on V-Sync, G-Sync, or neither?
For lowest input lag: V-Sync OFF, G-Sync ON, frame cap 3 fps below your monitor's refresh (e.g. 237 cap on 240 Hz). G-Sync alone holds frames to the refresh interval and adds 1–2 ms; V-Sync alone adds 16–50 ms by queuing frames. The combination 'G-Sync ON + V-Sync ON + capped below refresh' is the NVIDIA-recommended lowest-tear, lowest-lag setup. For competitive play where tearing is acceptable, all sync off plus uncapped FPS is fastest by about 1–3 ms. AMD FreeSync follows the same logic.
Is 1 ms GtG response time actually 1 ms?
Almost never. The '1 ms GtG' figure on the box is the manufacturer's cherry-picked best-case grey-to-grey transition with overdrive at maximum — often introducing inverse ghosting. The true average GtG across all colour transitions is typically 3–8 ms on a fast IPS, 1–3 ms on a quality OLED, and 5–15 ms on a VA panel. MPRT (motion-picture response time) is a different metric that includes pixel persistence and is dominated by refresh rate, not pixel speed. Trust independent reviewer numbers, not the sticker.