Ghosting test

This ghosting test moves shapes across dark, mid and light backgrounds at a set speed, in step with your screen's refresh rate. A trail behind them is ghosting, a bright or dark halo is overdrive overshoot, and a soft smear is motion blur.

Speed, px/s

Press Start the test, then follow one shape with your eyes as it crosses.

Refresh rate, measured
--
Move per refresh
--
Skipped refreshes
--

Motion

Ghosting, overshoot and blur

Match what you see around the shape to one of the four faults RTINGS names in its monitor motion tests.

  • Ghosting A faint trail behind the shape: pixels still changing from the last frame's color.
  • Overdrive overshoot A bright or dark halo the shape doesn't have: pixels pushed past their target shade. Reviews call it inverse ghosting.
  • Motion blur A soft smear on both edges, even on a fast panel. Most of it comes from your eyes tracking the shape while each frame holds still.
  • Dark smearing On a VA panel, a dark trail behind the black block on the mid and light bands, where pixels rise slowly out of black.

A shape that jumps instead of trailing has skipped a refresh, and the FPS test counts those as dropped frames.

Overdrive settings

Overdrive drives pixels harder to shorten the trail. DisplayNinja lists the names monitor menus use: Overdrive, OD, Response Time, or TraceFree on ASUS, with levels such as Normal, Fast and Faster. Too strong a level overshoots. RTINGS measured a 236% overshoot error on the ASUS XG27AQ at its strongest setting, on one transition to 20% gray.

The best level changes with the refresh rate, so tune it at the rate you play at, and confirm that rate with the refresh rate test. RTINGS also finds that a cold LCD responds more slowly, so let the screen warm up first.

Response

GtG and MPRT

GtG (gray to gray) is how long a pixel takes to change shade; ghosting and overshoot happen during that change. MPRT (moving picture response time) is how long each frame stays on screen; RTINGS notes that on a normal sample-and-hold screen that is the whole refresh.

A web page can't see the pixels change, so here slow GtG shows up as trails and halos, with no time in ms. Response time also decides how much you gain from a faster refresh rate, whether you're weighing 60Hz vs 120Hz or 144Hz vs 240Hz.

Why a 1 ms monitor still blurs

Your eyes follow a moving object while the screen holds each frame still, so the image smears across your retina. RTINGS explains it with the Blur Busters Law: each millisecond of persistence adds 1 px of blur for every 1,000 px/s of motion. In RTINGS' motion photos, this hold causes most of the blur, not slow pixels. At 960 px/s that is 16 px of blur on a 60 Hz screen and 6.7 px at 144 Hz, even with pixels that change instantly. Move per refresh, in the test above, is that blur for your own screen and speed.

A 1 ms GtG rating can't shorten that hold. A higher refresh rate or a strobe mode can, and the frame rate comparisons list the blur at each common rate. In RTINGS' strobing tests, backlight strobing and black frame insertion cut persistence, at the cost of a dimmer picture and possible flicker. Brands sell it under names such as ULMB 2 (NVIDIA), ELMB (ASUS) and PureXP (ViewSonic).

Method

How the ghosting test works

When you press Start, we time your screen's frames to measure its refresh rate. The shapes start moving on a first estimate after about a second. If your screen is faster than 60 Hz but the test finds 60, a setting, a cable or the browser is holding it back, and the refresh rate test names the likely cause.

  • Counted refreshes. The shapes move a set distance per refresh, not per unit of time, so once the rate is locked a skipped refresh still moves them the full distance.
  • Speeds. 960 px/s is the long-standing norm, and RTINGS now uses 1,920 px/s. They are in CSS pixels, the unit pages are laid out in.
  • Flashing. Each pass of a shape flashes a point 3 or 4 times, so we keep the shapes small. The area that changes from one frame to the next stays under a quarter of the 341 by 256 CSS px block that the W3C's WCAG 2.3.1 (Level A) measures flashes in. Narrow screens also lose the fastest speed, as an extra margin.
  • Backgrounds. Black, 50% gray and light gray, because RTINGS found that an overdrive level that helps dark scenes can overshoot in bright ones.
  • Your eyes, not a camera. A camera on a tripod does not record persistence blur, so reviewers such as RTINGS use a pursuit camera that slides along with the pattern.

Nothing moves until you press Start, and it stops when you switch tabs.

Questions

Can you give me an example of monitor ghosting?

In RTINGS' motion photos of the Samsung Odyssey G5, a VA monitor, a purple object on dark green leaves a dark trail. Red and blue switch off in 11.5 ms while green takes 27.4 ms to come on, so the pixels pass through near black. Overshoot looks different: the LG 32GR93U's Faster overdrive mode reaches its target within 1 ms but overshoots it, and RTINGS calls that mode unusable in practice.

How do I see if I have ghosting?

Press Start and follow one shape with your eyes across the dark, mid and light bands in turn, since a fault can show on one band only. Then match what trails the shape to the list under Ghosting, overshoot and blur.

How do I fix ghosting on my monitor?

Raise the monitor's overdrive one level at a time with this test running, and step back one level once a bright or dark halo appears. If the trail is gone but motion still looks soft, what is left is motion blur, which no overdrive level removes.

How can I check my monitor response time?

Look for a review that measures it: RTINGS puts a photodiode on the screen and times 72 gray-to-gray transitions at every overdrive setting. Go by the measured total, not the rating on the box, which is the maker's own figure.

What should the response time be on a monitor?

Shorter than one refresh, so each pixel settles before the next frame arrives. A 144 Hz monitor shows a new frame every 6.94 ms and a 240 Hz one every 4.17 ms, so those are the limits to beat. RTINGS' measurements show monitors with the same refresh rate differing widely on this, so the rate alone doesn't tell you.