Dead Pixel Warranty Rules: ISO 9241-307, Explained

Why "within spec" is a real sentence

A 1080p panel contains about 2.07 million pixels, and each pixel is three subpixels — one red, one green, one blue. That is over six million transistor-driven points of light, manufactured across a large sheet of glass, and producing every panel with all of them perfect would make displays dramatically more expensive. So the industry did what industries do with unavoidable imperfection: it standardized how much of it is acceptable.

The current framework is ISO 9241-307, which replaced the older ISO 13406-2. It defines the categories of pixel faults and the quality classes that set how many faults of each category a panel may ship with. A panel sold under a given class is warranted to that class — not to zero. When a support agent says your dead pixel is within spec, they are usually describing, accurately, the class your panel was sold at. The useful response is not to argue about fairness; it is to know the framework and measure what you actually have.

The three fault types

Type 1 — always on. The whole pixel is permanently lit: all three subpixels stuck on, producing a white dot. It is most visible on a solid black screen, where it is the only thing shining. Colloquially, a hot pixel.

Type 2 — always off. The whole pixel is permanently dark: a black dot on every color, most visible on a solid white screen. This is the dead pixel proper, and it is the least likely fault to ever recover.

Type 3 — the stuck subpixel. A defect in a single subpixel, which shows as a red, green, or blue point rather than a white or black one. You can identify which subpixel is stuck by cycling the primaries: a stuck-on red subpixel disappears into a full red screen and stands out against green and blue.

This classification is exactly what the color cycle in the dead pixel test is for. Wipe the screen first — a surprising share of reported dead pixels are dust — then step through black, white, red, green, and blue, and note which screens reveal each dot and what color it shows. Those two observations pin down the fault type, and the fault type is the vocabulary every policy is written in.

Fault classes and the per-million math

The standard's second half is the fault classes, which set how many faults of each type are permitted per million pixels. Class 0 permits none at all. Class I is the strictest tier you will commonly see sold, aimed at uses where any fault is unacceptable. Most consumer panels are sold as Class II, which tolerates a small number of faults of each type per million pixels. The exact per-class counts are the standard's fine print; what matters for a claim is which class your panel was sold under, and that Class II — the common case — does not promise zero.

The per-million basis has a consequence people miss: the allowance scales with resolution. A 4K UHD panel at 3840 × 2160 has 8,294,400 pixels — almost exactly four times the 2,073,600 of a 1080p panel — so at the same class it is permitted roughly four times as many faults. Higher resolution is not a stricter promise; it is a bigger denominator. The same three stuck subpixels that would exceed an allowance on a small low-resolution screen can sit comfortably inside it on a large 4K one.

What manufacturers layer on top

The ISO class is a framework, not the whole policy. On top of it, manufacturers write their own rules, and these differ in ways that decide real claims. Many policies distinguish bright dots from dark dots, tolerating fewer of the bright ones because a lit dot on a dark scene is far more visible than a dark dot on a bright one. Some premium product lines carry explicit zero-bright-dot promises — a guarantee that any always-on defect qualifies for exchange — while the standard tier of the same brand does not. Policies also commonly distinguish clustered faults from scattered ones, and replacement terms can differ between the initial return window and the longer panel warranty.

All of this varies by brand, by product line, and by region, and it changes over time. That is why this guide contains no brand-by-brand table: any such table starts aging the day it is written. The reliable move is to read the current pixel policy page for your specific model before you open a claim, and to bring measurements that speak that policy's language — fault type, count, and location.

Documenting a claim that holds up

First, clean the screen and re-check; dust removed is a claim you no longer need to file. Then run the dead pixel test across the full color cycle and write down what you find, in policy vocabulary: how many faults, what type each one is, and roughly where each sits — "one Type 3 stuck-on green subpixel near center, one Type 2 in the lower-left quadrant" is a report a reviewer can act on.

Photograph each fault on the background that shows it worst: bright dots on the black screen, dark dots on the white screen. Take one photo of the whole display and one close-up, and put an object of known size in the frame for scale — a ruler, or a bank card, which the ID-1 card standard fixes at 85.60 mm wide anywhere in the world. Note the purchase date, model, and serial number alongside the photos.

Timing matters more than most people expect. Retailer return rights inside the return window are often stronger and simpler than the panel warranty that follows — a return generally does not require you to argue fault classes at all. If you are inside that window, claim now rather than waiting to see whether the fault bothers you.

One exception is worth ten minutes before any paperwork: a Type 3 stuck subpixel sometimes recovers, and the fixer mode in the dead pixel test is a free, safe way to try. If the subpixel frees, the claim ends there. Type 1 and Type 2 faults almost never respond — for those, go straight to the documentation. Either way, the measurement is the point: it turns "my screen has a problem" into a typed, counted, photographed fault report, and that is the version of the conversation you can win.