Why Capping Frame Rates Below Display Refresh Ceilings Eliminates V-Sync Latency Under Variable Refresh Rate

Inside a darkened testing lab, a high-speed camera records a photodiode flashing against a gaming monitor at 1,000 frames per second. The measurements are ruthless. At 141 frames per second on a 144Hz display, the input lag registers at a blistering 3.47 milliseconds.[2]

But the moment the graphics card pushes that output to exactly 144 frames per second, the latency spikes by over 20 milliseconds. That sudden, invisible delay is the difference between landing a flick shot and missing it entirely. The culprit is a fundamental collision between Variable Refresh Rate technology and the legacy mechanics of V-Sync.

For years, competitive players treated V-Sync like a virus, disabling it universally to avoid the sluggish, heavy feeling it introduced to mouse movements. The arrival of Variable Refresh Rate protocols like NVIDIA's G-Sync and AMD's FreeSync promised a flawless compromise. These hardware modules allow the monitor to dynamically adjust its refresh rate to match the GPU's output.[3]

The Ceiling Collision

If the graphics card renders 112 frames in a second, the monitor refreshes exactly 112 times. The result is a tear-free image with zero added latency, fundamentally changing how displays handle motion. But that dynamic synchronization has a hard physical limit. A 144Hz monitor cannot refresh 145 times in a second.

When a powerful graphics card outpaces the display's maximum refresh rate, the VRR protocol runs out of headroom. It has nothing left to adjust. At that exact threshold, the system must decide what to do with the excess frames. If V-Sync is disabled, the monitor simply draws the new frame over the old one, resulting in aggressive screen tearing.

If V-Sync is enabled, the system reverts to its oldest habit, which is where the latency trap springs. Traditional V-Sync operates by forcing the graphics card to wait for the monitor's next refresh cycle before sending a new frame. When the GPU hits the refresh ceiling, it begins placing rendered frames into a queue.[4]

Bypassing the Buffer

This frame buffer ensures the image remains perfectly intact, but it means the frame you see on screen was actually rendered several milliseconds ago. According to display researchers at Blur Busters, this ceiling collision adds anywhere from two to four frames of absolute input lag.[1]

At 144Hz, a three-frame delay translates to roughly 21 milliseconds of added latency. In a competitive shooter where human reaction times are measured in fractions of a second, that buffer is an eternity. The player feels it as a subtle disconnect between their hand and the crosshair.

Because this penalty only triggers when the frame rate matches the refresh rate, a player's input lag will wildly fluctuate. It changes depending on whether they are looking at a complex, demanding scene or staring at a simple wall that allows their frame rate to spike. The solution to this fluctuating latency is a deliberate mathematical deficit.

The Hardware Handshake

By utilizing a frame rate limiter to cap the GPU's output slightly below the monitor's maximum refresh rate, the ceiling collision is entirely avoided. A cap of 141 frames per second on a 144Hz display ensures that the graphics card never produces a frame faster than the monitor can draw it.[5]

The VRR protocol remains active and in control 100 percent of the time, dynamically pacing the display without ever triggering the V-Sync queue. This specific configuration creates the optimal latency environment. The V-Sync setting remains active to handle any microscopic frametime variances and prevent tearing.

Because the frame rate never actually reaches the maximum refresh rate, the V-Sync buffer never engages. The player gets the visual perfection of V-Sync with the absolute minimum input lag of an uncapped system. The exact size of the required deficit depends on the precision of the frame rate limiter.[6]

Engine-Level Precision

In-game limiters built directly into the engine are generally the most efficient, applying the cap at the exact moment the frame is generated. This engine-level pacing introduces virtually zero additional latency, making it the preferred method for capping frame rates in competitive titles.

When an in-game limiter is unavailable, external limiters must step in. Tools like RivaTuner Statistics Server or the limiters built into GPU control panels operate at the driver level. They intercept the frames after they leave the engine but before they reach the display, which can add a microscopic amount of latency.

The standard recommendation from display engineers is a three-frame deficit, providing enough of a buffer to account for minor frametime spikes. Modern software has begun to automate this delicate balancing act. NVIDIA's Reflex technology is designed to minimize the render queue and reduce system latency dynamically.

The Role of Frametime Variance

When Reflex is enabled alongside G-Sync and V-Sync, it automatically applies a dynamic frame rate cap slightly below the monitor's maximum refresh rate. On a 240Hz display, Reflex will automatically cap the output at around 225 frames per second, ensuring the VRR ceiling is never breached.

For gamers utilizing AMD hardware, the mechanics of FreeSync operate under the exact same constraints. While FreeSync relies on an open standard rather than a proprietary hardware module, its behavior at the refresh ceiling is identical. If a Radeon graphics card pushes frames faster than the display can handle, the system will buffer.[3]

Applying a frame rate cap slightly below the maximum refresh rate is just as critical for Radeon users seeking the lowest possible input lag. The reason a three-frame deficit is recommended comes down to frametime variance. Even when a game is capped, the individual frames are not delivered at perfectly even intervals.

The Modern Standard

A sudden explosion or a rapid camera movement can cause a microscopic spike in rendering time. If the cap is too close to the ceiling, these minor variances can cause a single frame to briefly clip the maximum refresh rate, triggering a momentary V-Sync buffer.

By maintaining a slightly wider deficit, the system has enough breathing room to absorb these frametime spikes without ever touching the ceiling. The VRR protocol handles the fluctuating delivery times flawlessly, adjusting the monitor's refresh rate on the fly to match the uneven pacing.

Ultimately, the goal of any competitive display setup is consistency. By establishing a hard cap just below the monitor's physical limit, the latency remains flat and predictable. The hardware is allowed to deliver a flawless, tear-free image at the absolute speed of the player's inputs.[6]