How to Test a Monitor: Dead Pixels, Backlight Bleed and Color

A display is worth twenty minutes of testing on the day it arrives, for one unromantic reason: return windows close, and manufacturer dead-pixel policies are stricter than most buyers expect. Many warranties permit several dead subpixels before a panel qualifies as faulty, and a fault you notice in month three is your problem rather than the seller’s.

The tests below run in order from most to least likely to find something.

1. Dead and Stuck Pixels

The two defects sound alike and behave differently:

• Dead pixel — the transistor no longer receives power. It stays black on every background and cannot be recovered.
• Stuck pixel — one subpixel is permanently on, showing as a persistent red, green or blue dot. Sometimes recoverable.
• Hot pixel — all three subpixels stuck on, giving a bright white dot.

To find them, display full-screen solid colors — black, white, red, green, blue — and inspect the whole surface at each. A single black dot only shows against white and light colors; a stuck red subpixel only shows against black and blue. Testing on one color finds perhaps a third of the defects. A dead pixel test cycles the full set at true full-screen so nothing is hidden by browser chrome.

Two practical notes. Clean the screen first — dust and dried droplets impersonate stuck pixels convincingly. And a stuck pixel sometimes frees itself if the area is cycled rapidly through colors for 10–20 minutes, which is worth trying before concluding the panel is defective. Dead pixels never recover, because there is nothing left to stimulate.

2. Backlight Bleed and IPS Glow

Display a full black image in a dark room and look at the edges and corners. Two different things can appear:

• Backlight bleed — irregular bright patches, usually at the edges, where light escapes past the panel layers. It is a manufacturing defect, does not change with viewing angle, and is grounds for return when severe.
• IPS glow — a broad silver-gray haze in the corners that shifts or disappears as you move your head. It is inherent to IPS technology, present on every IPS panel to some degree, and not a defect.

The distinguishing test is simply to move: bleed stays put, glow moves. Judge both in a dark room on a genuinely black image, and give your eyes a minute to adapt — a display that looks perfect in daylight can reveal significant bleed at night, which is exactly when you will notice it in use.

3. Uniformity and Banding

Show a mid-gray full screen. Look for patches that are lighter, darker or tinted relative to the rest — cheap panels often show a visible vignette or a warm-to-cool gradient across the width. Then show a smooth gradient: banding appears as distinct steps instead of a continuous ramp, and indicates either a low-bit panel (6-bit with dithering rather than true 8-bit) or an aggressive processing mode.

Uniformity problems are hard to unsee once noticed, and they matter most for photo editing, design and any work with large flat color areas. For gaming and video they are usually invisible.

4. Color Accuracy

Without a hardware colorimeter you cannot measure accuracy, but you can catch gross errors:

• Grayscale ramp — steps from black to white should be neutral, with no green, magenta or blue tint creeping in at either end. A tint in the shadows is the most common factory error.
• Black level — a chart of near-black patches shows how many steps above pure black are distinguishable. If the first several steps look identical, the black level is crushed and shadow detail is being lost.
• White level — the same at the top end. Merged near-white patches mean highlights are clipped.
• Color ramps — each primary should transition smoothly from dark to saturated without shifting hue.

A color test set covers all four in sequence. Do the checks at the display’s native brightness in normal room lighting, not in the dark — a screen calibrated in darkness will look wrong every daytime hour you use it.

One setting worth correcting immediately: most displays ship in a "vivid" or "dynamic" preset that oversaturates color and pushes the white point toward blue, because it looks brighter in a shop. The standard, sRGB or neutral preset is almost always more accurate.

5. Refresh Rate and Frame Pacing

Confirm that the display is actually running at the refresh rate it advertises. A 144 Hz monitor connected with an inadequate cable, or left at defaults after a driver update, will happily run at 60 Hz without saying so. Measuring the real refresh rate in the browser shows what the system is delivering, which is the number that matters.

Common causes of an unexpectedly low rate: HDMI versions below 2.0 at high resolutions, cheap or long cables, the operating system’s display settings reverting after an update, and a laptop routing output through the integrated GPU rather than the discrete one.

Response Time and Ghosting

Quoted response times (often "1 ms") are measured under favorable gray-to-gray conditions and rarely describe the panel’s behavior across all transitions. What you can observe directly is ghosting — a trailing smear behind moving objects — and overshoot, sometimes called inverse ghosting, which appears as a bright halo when the overdrive setting is too aggressive. Most monitors offer several overdrive levels; the correct one is the highest setting that shows no visible halo. That is usually the middle option rather than the maximum.

6. Burn-In and Image Retention

These are different problems on different technologies. Image retention is temporary — a faint ghost of a static image that fades within minutes — and appears on both LCD and OLED. Burn-in is permanent, affects OLED, and results from uneven wear of the organic emitters where static elements such as taskbars and channel logos sit for thousands of hours.

To check a used OLED, display solid mid-gray and look for faint outlines of interface elements: a taskbar strip along one edge, a logo shape in a corner. A burn-in test cycles the appropriate colors, since some wear patterns are visible only on specific ones. On a modern OLED with pixel-shift and refresh routines, ordinary mixed use is generally fine; the risk sits with displays that show the same static interface all day.

7. Physical and Connection Checks

• Scratches and pressure marks. Inspect on both a black and a white screen at an angle. Pressure damage from shipping shows as a localised bright or discoloured area.
• Coating. Matte finishes reduce reflections and slightly soften text; glossy is sharper but mirrors the room. Neither is a defect — just check it suits your lighting.
• Every input. Test each HDMI and DisplayPort socket. A single failed port is easy to miss and impossible to claim later.
• Flicker. Some backlights dim using PWM, which causes headaches for sensitive users. Wave a hand quickly in front of a dimmed screen — visible stroboscopic "multiple hands" indicate PWM. Try it at several brightness levels, since many displays only use PWM below a threshold.
• Buzzing. Listen at low brightness with a static image in a quiet room. Coil whine is a legitimate return reason and does not improve with time.

Test Order and Conditions

1. Clean the screen.
2. Let the display warm up for 20–30 minutes; brightness and color drift while cold.
3. Set the picture preset to standard or sRGB, and turn off dynamic contrast and any eye-care or night mode.
4. Run pixel tests at full screen in a normally lit room.
5. Darken the room for backlight bleed and black uniformity.
6. Return to normal lighting for color, grayscale and gradient checks.
7. Verify refresh rate, then ghosting and overdrive.
8. Check every input and listen for buzzing.

Do all of this within the return window, and photograph anything questionable while the display is still returnable. A photograph of backlight bleed taken in a dark room is far more persuasive to a seller than a description of it.