Monitor Response Time vs Refresh Rate vs Input Lag: GtG, MPRT, Overdrive, and Latency Explained

A monitor can be described with several values measured in milliseconds, but they do not represent the same event. Pixel response time describes how quickly the panel changes from one luminance state to another. Refresh rate describes how often the display is able to refresh an image. Input or system latency describes delay across a wider chain that can include the input device, game engine, CPU, GPU, presentation queue, display electronics, scanout, and the visible pixel change.

That is why an advertised response-time value such as “1 ms” must not be read as “1 ms from mouse click to visible result.” NVIDIA’s Reflex measurement material explicitly treats system latency as the time from an input action to the resulting pixels changing on screen. Pixel response is only one component of that larger path, and the exact boundary of any latency number matters as much as the number itself.

Refresh rate is measured in hertz: refreshes per second. Its reciprocal gives the duration of one refresh interval. At 60 Hz one interval is about 16.67 ms; at 120 Hz it is about 8.33 ms; at 144 Hz about 6.94 ms; at 240 Hz about 4.17 ms; and at 360 Hz about 2.78 ms. Those values are deterministic timing relationships, not benchmark results.

A shorter refresh interval gives the display more frequent opportunities to present new image information and can reduce the display-side portion of latency, but it does not prove an equally small end-to-end latency. Game frame production, synchronization behavior, buffering, scanout timing, display processing, and pixel transitions still matter. NVIDIA’s 360 Hz material makes the same distinction by discussing faster scan time separately from measured system latency.

Gray-to-gray, usually written GtG or G2G, measures how long display hardware takes to transition between luminance levels. VESA’s Adaptive-Sync Display test specification defines G2G duration as the time required to move from one luminance level to another and uses a matrix of transitions rather than assuming one transition represents the whole panel.

This matters because LCD transitions are not all equally fast. A product can have a particularly quick best-case transition and slower transitions elsewhere. VESA’s certification work therefore evaluates multiple gray-to-gray transitions and also constrains overshoot and undershoot. A single manufacturer GtG figure is useful only when its measurement method and operating mode are understood; it is not automatically comparable with a number produced under a different method.

MPRT means moving-picture or motion-picture response time. Manufacturer explanations such as BenQ’s use it to describe how long image information remains visible during motion, which makes it a motion-blur or persistence-oriented concept rather than a direct replacement for a gray-to-gray transition measurement. Techniques such as backlight strobing can reduce visible persistence without turning the underlying LCD transition into the same numerical event.

VESA’s ClearMR program was created partly because time-only blur metrics such as MPRT can fail to capture image-quality penalties from blur-mitigation techniques. ClearMR instead evaluates the ratio of clear to blurry pixels and explicitly considers overshoot and undershoot. The practical conclusion is that GtG and MPRT answer different questions and should not be merged into one universal “monitor speed” score.

LCD overdrive applies a stronger drive signal to move a pixel toward its new state faster. VESA notes that overdrive is used to accelerate gray-to-gray transitions, but also measures overshoot and undershoot because excessive overdrive can create visible artifacts. NVIDIA’s latency guidance likewise recommends moderation and warns that too much overdrive can produce distracting artifacts that outweigh the response-time benefit.

Those artifacts are often described as overshoot, coronas, or inverse ghosting: instead of simply trailing behind a moving object, the pixel temporarily goes beyond the intended target and produces a bright or dark fringe. The correct setting is monitor-specific and can change with refresh rate or VRR behavior, so this guide does not prescribe a universal overdrive level. Product-level testing across several transitions and refresh rates is more useful than assuming the highest menu setting is fastest in practice.

Fast pixel transitions help a display reach the intended state before too much of the next refresh has elapsed, but transition speed is not the only source of perceived motion blur. Sample-and-hold displays keep each frame visible until the next refresh, so persistence also contributes to blur during eye tracking. Higher refresh rates shorten that persistence interval, while strobing or black-frame techniques can reduce visible persistence further by changing when light is emitted.

That is why two displays with the same advertised GtG value can look different in motion, and why a low MPRT claim does not automatically imply low overshoot, low signal-processing delay, or low total system latency. VESA’s separate Adaptive-Sync and ClearMR programs illustrate this measurement separation: gray-to-gray transition behavior, refresh behavior, blur, and artifacts are related but distinct properties.

A complete interaction can pass through the mouse or controller, application input handling, simulation, CPU submission, GPU rendering, synchronization and presentation, the display link, monitor processing, scanout, and finally the pixel transition that becomes visible. Measuring only one stage cannot establish the delay of the entire chain.

NVIDIA Reflex Latency Analyzer is a useful example of a broader measurement boundary because it detects an input event and measures until the resulting screen pixels change. That is fundamentally different from a panel GtG specification. When comparing latency claims, first ask where the timer starts and stops: device latency, render latency, display latency, and complete system latency are not interchangeable labels.

For refresh rate, convert hertz into refresh interval when you need the timing relationship, but do not claim that interval as complete input lag. For GtG, look for multi-transition measurements and overshoot behavior rather than one best-case transition. For motion clarity, consider persistence, strobing behavior, refresh rate, and a defined blur methodology. For latency, verify the measurement boundary and test mode.

A defensible monitor evaluation therefore asks several separate questions: how often can the image refresh, how quickly do pixels complete different transitions, how much overshoot accompanies overdrive, how clear is motion under the chosen mode, and how much delay exists from the relevant input or render event to visible output. No single “1 ms” label answers all of them, and Core Tech Tips does not convert one manufacturer response-time claim into an unsupported overall responsiveness verdict.