Ghosting test

Objects moving at a fixed real-world speed, so you can see trailing and overshoot and tune your overdrive setting against them.

Moving object

Background

Pauses first, then moves the object by a single frame so you can inspect an edge.

Speed

480 px / second

Position comes from elapsed time, so the object crosses the screen in the same number of seconds on any display.

Direction

Shortcuts

Space: Play or pause
H: Hide the controls
Esc: Leave full screen

What this test can and cannot tell you

The motion here is honest: position comes from elapsed time rather than a per-frame step, the object is moved on a compositor layer rather than through layout, and the speed in pixels per second is the speed you selected. That makes what you see comparable between two displays.

No millisecond figure comes out of this test, and none can. Pixel response is measured with a photodiode aimed at the panel, sampled far faster than any browser can observe. For measured grey-to-grey numbers, use a hardware review of your specific model.

For the rigorous version of this test, including the pursuit-camera method for photographing motion the way your eye sees it, go to TestUFO from Blur Busters, who developed the technique and have a decade of original research behind it.

More on this: what a browser cannot detect and how these measurements work.

How to run this test

Go full screen and press Play. The objects move at a fixed real-world speed, so the test behaves the same on a 60 Hz laptop and a 240 Hz monitor.
Track one object with your eyes as it crosses the screen, rather than staring at a fixed point. Following the object is what reveals trailing.
Raise the speed slider step by step and note where trailing first becomes objectionable. The three tracks let you compare the same speed at different heights on the panel.
Try the mid grey and black backgrounds. Dark transitions are slower on most panels, so trailing that is invisible on white often appears on black.
Change your monitor overdrive or response time setting one step at a time and repeat. This is the setting the test exists to help you tune.

What a problem looks like

Ghosting: A soft, darker trail following the moving object, caused by pixels taking time to finish changing. It is worst on dark-to-light and light-to-dark transitions and gets more obvious as speed increases.
Inverse ghosting, or overshoot: A bright halo or corona in front of or behind the object, often with a distinctly different colour to the background. This is overdrive pushing the pixel past its target value and then settling back. It means your overdrive setting is too aggressive, and turning it down one step usually fixes it.
Black smearing: A heavy dark smudge trailing an object over a dark background, far worse than the trailing on white. This is characteristic of VA panels, whose dark transitions are slow. Changing overdrive helps only a little; it is largely inherent to the panel type.
Blur with no visible trail: If the object looks soft while moving but leaves no distinct trail behind it, you are seeing persistence blur rather than ghosting, and the cause is how long each frame is held on screen. The motion blur test covers that.
Stutter rather than trailing: If the object jumps rather than sliding, the problem is frame delivery, not pixel response. Check the refresh rate test for dropped frames before adjusting any display setting.

What to do next

If something looks wrong, check it again in a normally lit room and from straight on, then check whether it appears on a second device. If it only happens on this screen and you are still inside the return window, that is the time to contact the retailer.

Questions people actually ask

What is the difference between ghosting and inverse ghosting?

Ghosting is a dark trail caused by pixels changing too slowly. Inverse ghosting is a bright halo caused by overdrive overshooting the target value in an attempt to change them faster. They have opposite causes and opposite fixes: more overdrive for one, less for the other.

How do I set overdrive correctly?

Raise it one step at a time while watching this test, and stop at the last setting before bright halos appear. The strongest setting is almost never the best one. If your monitor labels the options rather than numbering them, work from the mildest upwards.

Can this tell me my response time in milliseconds?

No. Measuring grey-to-grey or moving picture response time needs a photodiode and an oscilloscope pointed at the panel, which is what hardware review sites use. A browser cannot do it, and any site that prints a millisecond figure from a web page is not measuring your panel.

How do I capture what I am seeing?

With a pursuit camera: a camera that pans alongside the moving object at the same speed, so it records what your eye sees rather than what a static camera sees. Blur Busters documented the technique and their TestUFO site has patterns built for it, which is the right place to go for rigorous measurement.

Why does the test look the same at 60 Hz and 144 Hz?

The object speed is derived from elapsed time rather than from a step per frame, so it crosses the screen in the same number of seconds on any display. That is deliberate: it makes the amount of trailing comparable between machines instead of scaling with refresh rate.