Monitor Response Time: GtG vs MPRT Explained
GtG and MPRT are not the same measurement, so a monitor advertised at 1ms MPRT is not a 1ms GtG monitor. GtG (grey-to-grey) times how long a pixel takes to change from one shade to another. MPRT (moving picture response time) times how long a frame remains visible to a moving eye, and on a conventional sample-and-hold display that is at minimum the frame time - about 4.2ms at 240Hz - however quickly the pixels themselves switch. Persistence, not pixel speed, causes most of the blur people actually see.
Check price on Amazon →Part of the gaming monitor guide. The underlying display physics is covered in how display panels actually work.
What GtG actually measures
GtG, or grey-to-grey, is the time a single pixel takes to travel from one grey level to another. A measuring instrument watches one point on the panel, records the luminance curve as the pixel changes, and reports the duration of that transition. Most figures exclude the first and last 10 percent of the change, because the tails of the curve are slow and including them would produce far larger numbers.
The GtG figure printed on a monitor box is a best case. A panel has hundreds of possible transitions, and they are not equally fast. Dark-to-dark transitions on VA panels are notoriously the slowest, which is why a VA monitor rated 1ms GtG can still leave a visible dark smear behind moving objects on a dark background. The advertised number is usually the single fastest transition the panel can perform, at its most aggressive overdrive setting, at room temperature.
GtG matters, and a slow GtG panel produces a distinct artefact: a coloured or dark trail attached to the trailing edge of a moving object. But GtG is a floor, not a ceiling. Once GtG is fast enough to complete well inside a refresh interval, making it faster stops changing what the eye sees.
What MPRT measures, and why it is a different quantity
MPRT, or moving picture response time, measures how long a frame is visible to a moving eye rather than how fast a pixel changes. It is a measurement of persistence: the duration for which the display holds a static image in front of a viewer whose gaze is sweeping across the screen.
Conventional LCD and OLED monitors are sample-and-hold displays. Each frame is drawn and then held, unchanged and continuously lit, until the next frame replaces it. That hold is the entire refresh interval. At 60Hz the frame is held for roughly 16.7ms, at 144Hz for roughly 6.9ms, at 240Hz for roughly 4.2ms, at 480Hz for roughly 2.1ms. Those numbers are the MPRT floor for a non-strobed display at that refresh rate. No improvement to pixel transition speed can push MPRT below that floor, though slow transitions can push it above.
This is why the two specifications can both be true at once and still describe very different displays. A monitor can honestly claim 4ms GtG and 1ms MPRT, because the 1ms MPRT figure is measured with a strobing or black-frame-insertion mode active, not in the mode most owners run day to day.
Why a fast panel still looks blurry: eye tracking and sample-and-hold
The blur is generated in the eye, not in the panel. When a viewer tracks a moving object across the screen, the eye moves smoothly and continuously. The display does not. The display shows a frame, holds it stationary for the whole refresh interval, then jumps the object to its next position.
During that hold, the eye keeps sweeping while the image stays still. The stationary image is therefore smeared across the retina for the full duration of the hold. The result is a blur whose length depends on how far the eye travelled during the hold - which is a function of persistence and object speed, and nothing else.
This is the reason OLED monitors, whose pixel transitions complete in a fraction of a millisecond, still show motion blur at 120Hz. The pixels are effectively instant. The hold is not. It also explains why old CRTs looked sharper in motion than modern panels with vastly better response figures: a CRT illuminated each phosphor for a very short flash and then went dark, so persistence was low even at 60Hz.
Converting persistence into pixels of blur
Blur Busters expresses the relationship as a simple rule: 1ms of persistence produces 1 pixel of motion blur for every 1000 pixels per second of motion. The arithmetic is direct, and it makes the specification concrete.
Take a 240Hz sample-and-hold monitor, MPRT roughly 4.2ms, and an object crossing the screen at 1000 pixels per second - a slow pan. The blur trail is about 4 pixels wide. At 60Hz the same pan blurs across roughly 17 pixels. Speed the motion up to 3000 pixels per second, which is a normal flick or a fast strafe, and the 240Hz display smears that object across roughly 13 pixels. Even 1ms MPRT, the number monitors advertise, still produces about 3 pixels of blur at 3000 pixels per second.
The practical consequence is that refresh rate, not GtG, is the main lever on perceived motion blur for any display that is not strobing. Halving the frame time halves the blur, provided the frame rate rises with the refresh rate - a frame held on screen for 16.7ms blurs like 60Hz whatever the panel's maximum. Shaving GtG from 4ms to 1ms on a 144Hz panel changes the trailing smear but leaves the underlying persistence blur untouched.
What overdrive does, and how it overshoots into inverse ghosting
Overdrive is the panel's method of forcing pixels through their transitions faster. Instead of driving the liquid crystal to the voltage that corresponds to the target shade, the controller briefly applies a stronger voltage - overshooting the target - so the crystal accelerates, then pulls back to the correct level once the pixel is near its destination. Done well, this cuts GtG substantially with no visible cost.
Done too aggressively, the pixel arrives past its target before the voltage is pulled back. A pixel travelling from mid grey to light grey briefly goes brighter than light grey; a pixel travelling downward briefly goes darker than intended. On screen this appears as a bright halo or a pale outline leading or trailing a moving object. That artefact is overshoot, and the visible result is called inverse ghosting - inverse because a conventional ghost is a dark trail left behind, whereas this one is a bright edge.
The compensation itself is tuned per transition. The scaler holds a lookup table indexed by the pixel's starting grey level and its destination grey level, so every transition pair has its own drive value; manufacturers typically store a coarse table, commonly 17 by 17, and interpolate it up to the full 256 by 256 grid to save scaler memory. What is global is the user-facing control. The monitor ships with a handful of pre-baked tables and the Normal, Fast or Extreme setting simply selects which one applies to the whole panel. That is why the highest overdrive mode usually looks worse than the middle setting: the aggressive table raises compensation across the board, fixing the slowest transitions at the cost of pushing transitions that were already fast into overshoot. The practical advice is to test the middle overdrive mode against the highest one on a moving test pattern and choose the one with no bright fringing, not the one with the highest number.
Overdrive tuning is also refresh-rate dependent. A setting that is correct at 240Hz is frequently too aggressive at 60Hz, because the pixel has far longer to overshoot before the next frame arrives. Monitors with variable overdrive swap tables automatically as the frame rate changes; monitors without it will show fringing at some point in the VRR range.
Black frame insertion and backlight strobing: buying clarity with brightness
Since persistence blur comes from holding an image lit for the whole refresh interval, the fix is to stop holding it lit. Two implementations do this. Black frame insertion inserts a black frame, or blanks the panel, between displayed frames. Backlight strobing pulses an LCD backlight on for a short window and keeps it off the rest of the cycle. OLED displays, which have no backlight, must use the black-frame approach. Manufacturer names include NVIDIA ULMB and ULMB 2, BenQ DyAc, ASUS ELMB and versions marketed simply as 1ms MPRT.
The effect is large. If a 240Hz monitor lights each frame for only 1ms instead of 4.2ms, MPRT drops to 1ms and blur drops by the same factor. NVIDIA quotes a 360Hz display running ULMB 2 as delivering roughly 1440Hz of effective motion clarity, and requires certified monitors to strobe at the full refresh rate, reach at least 250 nits while strobing, and show minimal crosstalk.
The cost is light. The panel is dark for most of each cycle, so average brightness typically falls by roughly a third to a half, and considerably more at the shortest strobe pulse lengths. The 1440Hz figure above is simply the arithmetic of a 25 percent duty cycle at 360Hz, and the most aggressive strobe settings run shorter windows still. Measured losses are milder than the bare duty cycle implies, because monitors drive the backlight harder during the pulse to claw some of it back. Shorter pulse, less blur, dimmer picture. That trade is unavoidable: the two quantities are the same physical variable. Strobing also reintroduces flicker at the refresh rate, which some viewers perceive as eye strain, and it is more noticeable at lower refresh rates.
There is a second artefact to watch for. Strobe crosstalk appears as a faint double image, usually near the top or bottom of the screen, and it happens when the backlight fires while some rows of the panel are still finishing their transitions. Newer implementations attack this by tuning overdrive per screen region - NVIDIA calls this Vertical Dependent Overdrive - so that the rows the strobe illuminates have already settled. Many strobing modes also cannot run at the same time as variable refresh rate, so enabling one usually means giving up the other.
Which number should influence the purchase
Read the advertised response time as an advertisement, not a measurement. If a monitor claims 1ms without saying GtG or MPRT, assume the more flattering of the two. If it claims 1ms MPRT, assume that figure requires a strobing mode that most owners will leave off because of the brightness penalty.
For motion clarity in normal use, refresh rate is the specification that determines the outcome, because it sets the persistence floor. GtG needs only to be fast enough to finish inside a refresh interval; beyond that, further reductions change nothing the eye can see. On a 240Hz panel the refresh interval is 4.2ms, so any panel genuinely completing most transitions in one or two milliseconds is adequate, and the remaining blur is persistence.
The one case where GtG still decides the purchase is a panel with badly uneven transitions. A VA panel with slow dark transitions will smear in dark scenes regardless of refresh rate, and no overdrive preset fully fixes it without introducing overshoot elsewhere. This is where independent transition-by-transition measurements are worth consulting, because a single averaged GtG number hides exactly the failure that will annoy you.
Related
- 1440p vs 4K Gaming Monitor: 27in and 32in Compared
- IPS vs VA vs OLED for Gaming: Which Panel Wins?
- Mini-LED vs OLED Monitor: Zones, Blooming, Brightness
- Is a Higher Refresh Rate Worth It? 144Hz to 480Hz Explained
Worth Buying Alongside
Monitor arm (VESA gas-spring)
Frees desk depth and lets you set height and viewing distance precisely, which affects eye strain more than any panel spec.
Shop on Amazon →DisplayPort 2.1 cable
Needed for the highest bandwidth modes. An older cable silently caps refresh rate or forces compression.
Shop on Amazon →HDMI 2.1 48Gbps cable
Required for 4K/120 from a console. Unlabelled "high speed" cables are not guaranteed to carry it.
Shop on Amazon →USB-C 90W video cable
For single-cable laptop docking on monitors that offer power delivery.
Shop on Amazon →Monitor bias lighting
Raises perceived contrast in a dark room and reduces eye fatigue without touching picture settings.
Shop on Amazon →Microfibre screen cloth
OLED coatings scratch easily, and household cleaners can damage them.
Shop on Amazon →Frequently Asked Questions
No. They measure different things and are not interchangeable. 1ms GtG means a pixel completes a grey-to-grey transition in one millisecond. 1ms MPRT means each frame is visible to the eye for only one millisecond, which on a sample-and-hold display requires strobing or black frame insertion, since the natural persistence at 240Hz is about 4.2ms. A 1ms MPRT rating usually says more about a strobing mode than about panel speed, and the same monitor might measure 4ms or worse on GtG.
Because most visible motion blur comes from persistence, not pixel transition speed. A conventional monitor holds each frame lit for the entire refresh interval while your eye keeps moving to track the object, smearing that held image across the retina. At 144Hz the hold is about 6.9ms, which produces roughly 7 pixels of blur for every 1000 pixels per second of motion, regardless of how fast the pixels switch. Raising refresh rate or enabling a strobing mode reduces this; a faster GtG figure does not.
Overshoot happens when a monitor's overdrive applies too much voltage to accelerate a pixel transition, so the pixel goes past its target shade before settling back. Inverse ghosting is what overshoot looks like on screen: a bright halo or pale outline around moving objects, as opposed to the dark trailing smear caused by slow pixels. It is most common on the highest overdrive setting, often labelled Extreme or Fastest. The fix is to lower the overdrive setting until the bright fringing disappears.
Usually the middle one. Overdrive compensation is tuned per transition inside a lookup table indexed by starting and destination grey level, but the monitor exposes only a few pre-baked tables and your setting picks one for the entire panel. The most aggressive table raises compensation across the board, so it fixes the slowest transitions while pushing already-fast ones into overshoot. Display a moving test pattern, cycle through the settings, and pick the highest one that shows no bright fringing or pale outline behind moving objects. The correct choice also changes with refresh rate - a mode that is clean at 240Hz can overshoot badly at 60Hz - so monitors with variable overdrive handle VRR gameplay better.
Both, and the blur reduction is the larger visual change. On a sample-and-hold display, persistence equals the frame time, so doubling refresh rate halves the frame time and halves the blur. Moving from 60Hz to 240Hz cuts persistence from about 16.7ms to about 4.2ms, which shortens the blur trail by a factor of four. This holds only if the frame rate actually rises with the refresh rate; running 60fps on a 240Hz display leaves each frame on screen for 16.7ms and reproduces 60Hz blur.
Yes, though not from pixel response. OLED pixel transitions complete in a fraction of a millisecond, so trailing smear is essentially absent and there is no overshoot to tune. But an OLED is still a sample-and-hold display: it holds each frame lit for the full refresh interval, so its persistence blur matches that of an LCD with perfect transitions at the same refresh rate. Real LCDs add GtG smear on top of that, which is why an OLED still shows less total motion blur than a typical LCD at the same refresh rate. Reducing the persistence component requires either a higher refresh rate or black frame insertion.
Brightness, primarily. The panel is dark for most of each cycle, so average output typically falls by roughly a third to a half, and considerably more at the shortest strobe pulse lengths - a shorter lit window means less blur but a dimmer image, because the two are the same variable. Strobing also introduces flicker at the refresh rate, which some people find fatiguing, especially below 120Hz. It can produce a faint double image called strobe crosstalk, and on many monitors it cannot run at the same time as variable refresh rate.
Strobe crosstalk is a faint double image, usually visible near the top or bottom of the screen, that appears when backlight strobing is enabled. It occurs because an LCD refreshes row by row from top to bottom, so when the backlight fires, some rows may still be completing their transitions and are briefly illuminated mid-change. Implementations reduce it by timing the strobe to the window when most rows have settled, and by tuning overdrive differently for different screen regions - NVIDIA's term for the latter is Vertical Dependent Overdrive.
Treat it as a best case rather than a description of typical behaviour. Advertised GtG figures generally reflect the single fastest transition the panel can perform, at maximum overdrive, and exclude the slow tails of the transition curve. Advertised MPRT figures generally require a strobing mode that costs substantial brightness. Neither is fraudulent, but neither describes what the monitor does in a normal game at default settings. Independent transition-by-transition measurements are far more informative, particularly for VA panels where dark transitions are the weak point.