Greyscale Test

A greyscale test (also spelled grayscale in the US) shows neutral grey fields at known luminance values so you can judge whether your display reproduces grey as grey. Grey contains no hue, which makes it the cleanest possible reference: any colour or brightness error in the display chain becomes visible against it. This test gives you six fields covering the three faults that grey exposes. The 5% gray field and the 32-step ramp reveal black crush, where the darkest steps collapse into flat black. The 50% neutral gray field and the 5-step ladder reveal colour cast and gamma errors. The 5x5 and 4x4 uniformity grids reveal backlight bleed and OLED mura. Together they take about three minutes.

Set brightness to your normal working level and disable dynamic contrast, eco mode, and any ambient light sensor, then clean the screen. Click Start Test, which opens the fullscreen 5% gray field; look at the corners from your normal seating distance and note whether the dark grey is visible at all. Press the right arrow to move through the 5-step ladder and confirm all five patches differ. Continue to the 32-step ramp and count how many steps you can separate, paying attention to the darkest third. Finish with the 5x5 and dark uniformity grids, looking for patches brighter or dimmer than their neighbours. The whole sequence takes about three minutes, and the diagnostic panel scores the result as you go.

The 5 gray test is the common name for checking the five anchor grey levels that matter most: 5%, 10%, 25%, 50%, and 75%. Of these, 5% gray is the most demanding, because it sits roughly 13 levels above pure black on an 8-bit scale and is the first value to disappear when gamma is set too high or when the RGB range is mismatched. On a correctly configured display, a 5% field on a black background looks like a clearly visible dark grey, not a black rectangle. If you cannot see it, shadow detail in games, dark films, and photo editing is being thrown away before it reaches your eyes, and no amount of brightness will bring it back.

Black crush has three common causes, and they are worth separating because only one of them is a hardware problem. The most frequent is a gamma setting that is too high, typically a picture preset labelled Dynamic or a gamma value above 2.2, which steepens the curve near black. The second is an RGB range mismatch: if the graphics driver sends Limited range while the display expects Full, or the reverse, the lowest values are clipped. The third is aggressive black adjustment, such as dynamic contrast or an OLED black equaliser, which deliberately discards near-black steps. Check gamma and RGB range first, since both are free to fix, and only then suspect the panel itself.

Grey uniformity describes how consistent the same grey value looks across the whole screen area. To measure it visually, the test fills the panel with a grid of identical 50% grey patches separated by thin dark lines, and you compare each patch with its neighbours. A uniform panel shows twenty-five identical squares. Backlight bleed appears as one or more corners that are brighter than the rest, clouding appears as a smudged bright region in the middle, and edge dimming is a gradual falloff towards the borders. The 4x4 dark grid does the same job at 5% grey, where OLED mura and LCD clouding are at their most visible. Judge both from normal viewing distance, not up close.

It depends on severity, and most manufacturers publish a threshold rather than accepting any variation. A reasonable test is to photograph the 5x5 grid from your normal viewing distance in a normally lit room, not in a dark room with a long exposure, because long exposures exaggerate the fault and service centres reject those images. If a patch or corner is obvious in that photo and remains obvious after a factory reset at moderate brightness, you have a case worth opening. Note that edge dimming is often considered normal on curved and thin-bezel panels, and that most warranty terms exclude a small number of bright subpixels. Document the fault within the return window, since policies are stricter after it closes.

OLED needs a slightly different procedure because it has no backlight. Open the 5% gray field in a dark room and let the panel warm up for about fifteen minutes first, since OLED luminance shifts with temperature. What you are looking for is mura: faint vertical bands or blotches that appear only at very low grey levels, plus any pink or green tint affecting part of the screen. Small variations are normal even on premium sets. Check the dark uniformity grid at normal viewing distance rather than up close, because at 30 centimetres every OLED shows structure. Do not leave a static grey field on screen for long periods, as it is exactly the kind of pattern that accelerates burn-in.

Partly, and it is worth knowing the limits. A 50% neutral grey field is an excellent visual check for colour cast: if the grey looks blue, green, pink, or yellow rather than neutral, your white point is off, and most displays let you correct it with an RGB gain control or a preset such as Warm or Standard. The 32-step ramp gives a practical read on gamma, because too high a gamma hides the darkest steps while too low a gamma washes out the lightest ones. What a browser test cannot do is report numbers. It cannot tell you that your gamma is 2.34 or your white point is 6800K, because that requires a colorimeter and calibration software.

It is accurate for everything that depends on the pattern itself, and it is completely free with no signup. The grey values are hard-coded and rendered locally, so the 5% field really is 5% and the uniformity patches really are identical, which is what makes visual comparison valid. What no browser test can do is measure absolute luminance in nits or verify colour volume, because the browser has no access to a hardware sensor. In practice that is a good division of labour: use this test to find black crush, colour cast, and uneven panels in three minutes, and use a colorimeter only for work where numerical accuracy in nits and colour space coverage actually changes the outcome.

A 5% gray field on a black background is a demanding stress test, and barely seeing it is common on cheap laptop panels because their native contrast ratio is low. A good LCD should still show the 5% field as a clearly distinguishable dark grey at a normal viewing brightness, and a correctly configured OLED usually reveals it more easily because its black level is far deeper, which widens the contrast between that step and the background behind it. When the field is invisible or blending into pure black, the panel is crushing that step. Check your black level and RGB range, and compare the same field across two displays by running the 5% gray pattern on each. The tool also keeps 5% grey available inside the 32-step ramp so you can confirm the reading twice.

Grey uniformity describes how consistent the same grey level looks across the whole surface, while backlight bleed is a specific cause of an uneven result in an LCD. Bleed is the light that leaks from the edges of the backlight assembly around the panel and shows up as brighter corners or a bright band along a bezel, usually on a 5% grey field rather than a mid grey. During the 5x5 test, a uniform panel shows 25 identical squares, whereas bleed in a corner makes the nearby patches brighter. OLED panels have no backlight and therefore no bleed at all, so a cloudy bright patch on an OLED comes from the panel itself, not from light leakage. Bleed is a property of the backlight assembly, while uniformity describes the panel quality you are actually assessing.

Yes, for the three faults that genuinely justify a return: black crush, colour cast, and dark uniformity. Run the 5% field in a dark room and the 4x4 dark grid at your normal viewing distance. If a mura patch is obvious from your normal viewing seat, or a colour cast that persists across the 50% grey field in all picture presets, you have a case for a return. Before contacting the retailer, reset to factory picture settings, set gamma to 2.2, and take a photo of the grid from the normal distance, not a long exposure that magnifies the fault. Most retailers replace panels with visible mura from an acceptable angle; a very small variation of about 5% is usually judged normal.