"No dead pixels" sounds like a clear purchasing requirement until a shipment is inspected. Is one permanently bright subpixel a failed pixel? Does a defect in the center of a menu board matter more than one at the edge? Are two adjacent defects worse than two defects on opposite corners? Which test pattern, viewing condition, and evidence count when the supplier and buyer disagree?
A commercial LCD project needs a written acceptance method, not consumer-display folklore. The method should define the exact assembly being inspected, the defect terminology used in the contract, the test condition, any location or clustering rules, the quantity limits supplied by the applicable specification, and the disposition process for units that fail. This article shows how to build that method without inventing a universal pixel-defect count.

Start with definitions, because "dead" and "stuck" are not contract language
An LCD picture element is built from subpixels, commonly color subpixels that combine to form the visible pixel. In everyday usage, "dead pixel" may mean a pixel that remains dark, while "stuck pixel" may mean a pixel or subpixel that remains on or shows a constant color. Supplier documentation can use different categories, so the inspection plan should copy the terminology and definitions from the exact production model or contract rather than assuming everyone uses the words the same way.
At minimum, the acceptance record should distinguish:
- a full pixel defect from a subpixel defect;
- a bright or stuck-on defect from a dark or stuck-off defect, if the supplier uses those categories;
- isolated defects from adjacent or clustered defects;
- defects inside a project-defined critical viewing zone from defects outside it, when location matters;
- panel pixel defects from artifacts caused by signal, scaling, content, protective glass, contamination, or image retention.

Know what the ISO documents do-and what they do not do
ISO 9241-307:2008 remains current after ISO's 2024 review and establishes analysis and compliance test methods for electronic visual displays. Its lifecycle page also identifies the older ISO 13406-2:2001 as withdrawn. Buyers should therefore be cautious about copying a pixel-count table from an old web page and labeling it a current ISO requirement without checking the applicable document and product contract.
ISO/TR 9241-310:2010, confirmed in 2022, is specifically about the visibility, aesthetics, and ergonomics of pixel defects. ISO explicitly states that this technical report provides guidance but does not itself give requirements related to pixel defects.
The procurement consequence is simple: standards can provide test and terminology context, but the project still needs model-specific acceptance criteria. Ask the supplier to identify the exact specification, class, panel policy, warranty document, or contractual limit that applies to the production unit being purchased.
Build an inspection station that can reproduce a dispute
Pixel inspection should make a defect easier to classify, not easier to argue about. A practical receiving or factory inspection method should define the following conditions before the first batch arrives.
- Use the production display assembly. Record the display model, panel or assembly revision where available, controller/player, and relevant firmware or image-processing settings.
- Drive the panel at its intended native mapping. Avoid introducing scaling artifacts that can imitate a pixel problem. Where the product provides a diagnostic or native test mode, use the approved method.
- Use a controlled pattern set. Solid black, white, red, green, blue, and suitable gray patterns can reveal different defect types. The exact set should be documented so supplier and buyer test the same condition.
- Define the viewing condition. Record ambient lighting, inspection distance, display orientation, and any other condition that materially affects visibility. A close microscope-style inspection and a normal-use inspection answer different questions.
- Stabilize the display as required. Follow the applicable manufacturer procedure before inspection rather than mixing cold-start observations with stabilized operation.
- Record location and evidence. Note the screen coordinates or a repeatable grid location, defect category, pattern on which it appears, photograph if useful, and the disposition decision.
If a defect cannot be reproduced on the same test patterns and mapping, investigate the source path before classifying the panel.
Use a defect map instead of a single total count
A total count can hide the characteristics that actually drive visual impact. A better acceptance sheet has separate fields for defect type, location, and clustering.
| Field | What to record | Why it changes the decision |
|---|---|---|
| Defect category | Supplier-defined bright, dark, subpixel, full-pixel, or other applicable category | Different categories may have different permitted limits. |
| Screen location | Coordinate or inspection grid | A project may apply a stricter rule in a central or content-critical zone. |
| Adjacency / cluster | Whether defects are isolated or near one another | Several nearby defects can be more noticeable than the same number distributed across the screen. |
| Pattern visibility | Which approved test colors expose the defect | Helps distinguish defect type and reproduce the observation. |
| Normal-use visibility | Whether the defect is visible from the defined operating position | Useful where the contract includes an ergonomic or aesthetic viewing criterion. |
| Evidence | Photo, inspection record, unit serial/asset ID, batch, inspector | Turns a subjective complaint into a traceable supplier disposition. |
Define critical zones and cluster rules before inspection starts
Some projects care about where a defect appears. A price, logo, face, menu item, or small-font information area may make a central defect operationally more noticeable than one at an unused border. If location matters, draw the critical zone in the specification before seeing the shipment. Do not move the boundary after a result is known.
The same principle applies to clusters. Define what counts as adjacent or clustered using the supplier or project method. A statement such as "no clusters" is still ambiguous unless the contract explains the distance or geometry used to classify one.
For ultra-wide displays, consider whether the content layout itself creates multiple critical areas rather than assuming a conventional central rectangle. The visual-use case should drive the zone definition.

Turn supplier acceptance criteria into a one-page contract table
Do not ask only "What is your dead-pixel policy?" Ask for the exact production-model rule in a format that can be used during receiving inspection and warranty review.
| Requirement | Project entry |
|---|---|
| Production display / panel identification | Exact approved model and revision |
| Referenced standard or supplier policy | Document name, revision, date, and applicable class or category if used |
| Bright / stuck-on defect limit | Enter the model-specific contractual limit; do not copy a generic number |
| Dark / stuck-off defect limit | Enter the model-specific contractual limit |
| Subpixel treatment | State how subpixel defects are counted |
| Cluster rule | Define adjacency and allowed quantity |
| Critical-zone rule | Attach the zone drawing and any stricter limit |
| Inspection condition | Patterns, mapping, ambient condition, viewing distance, stabilization |
| Failure disposition | Rework, replace, accept by concession, or other approved route |
| Warranty / field rule | State whether the same criteria apply after installation and over what contractual period |
Do not misclassify a source or surface artifact as a pixel defect
Before opening an RMA, determine whether the apparent point follows the panel or the signal path.
- Content or rendering artifact: a mark that moves with the image is not a fixed panel pixel defect.
- Scaling or mapping issue: non-native scaling can create fine-line or single-pixel-looking artifacts that disappear with a native diagnostic pattern.
- Cable or interface fault: intermittent digital errors usually do not behave like one permanently fixed pixel. Check the source path when the symptom changes with cable, player, or port.
- Surface contamination: dust, protective film, adhesive residue, or debris on or under protective glazing can imitate a dark point. Use the approved cleaning and inspection method before diagnosis.
- Image retention: persistent content is a different mechanism and should be evaluated separately. The LCD Image Retention Prevention guide covers that operational risk.

Use sampling for process control, but preserve unit-level disposition
A buyer may use a sampling plan for incoming quality control, especially for large orders, but the sampling method and the pixel-defect acceptance criteria are different decisions. The sampling plan determines how many units are inspected and how a lot is accepted or escalated; the pixel policy determines whether an inspected unit passes.
Define both in the purchase quality plan. If the project uses an AQL or another sampling framework, the sampling values should come from the buyer's approved quality system and contract-not from this article or from an unrelated display supplier.
Make the RMA evidence hard to dispute
For each failed unit, capture the asset or serial identifier, panel/display model, test pattern, defect map, image or video evidence where useful, inspection condition, applicable contractual rule, and disposition. If the supplier requests a retest, reproduce the agreed method rather than changing the viewing distance or pattern ad hoc.
For commercial product context, see the Bar LCD Display Screen collection. Luminance aging and replacement decisions are a separate issue covered in Commercial LCD Backlight Luminance Drift Monitoring.
Questions buyers should ask before approving the supplier
- What exact pixel-defect policy applies to this production display or panel?
- Which standard, document revision, class, or internal specification defines the acceptance limits?
- How are full-pixel and subpixel defects counted?
- Are bright and dark defects treated differently?
- Is there a cluster or adjacency rule?
- Does the contract distinguish central or critical viewing zones?
- What test patterns, native mapping, viewing distance, ambient condition, and warm-up procedure are required?
- Does the same policy apply at incoming inspection and during the warranty period?
- What evidence is required for replacement, rework, or RMA authorization?
FAQ
Q: Does ISO 9241-310 specify a universal number of allowed dead pixels?
A: No. ISO/TR 9241-310 provides guidance on visibility, aesthetics, and specification of pixel defects, and ISO explicitly says the report does not itself give pixel-defect requirements.
Q: Is ISO 13406-2 still the current ISO reference for flat-panel pixel defects?
A: ISO lists ISO 13406-2:2001 as withdrawn in the lifecycle of ISO 9241-307:2008. Buyers should verify the current applicable standard and the supplier's exact contractual policy instead of copying an old class table from secondary websites.
Q: Should a central defect automatically fail a commercial display?
A: Only if the applicable supplier policy or project contract says so. If location sensitivity matters, define the critical zone and its rule before inspection.
Q: Can a stuck-looking point be caused by the video source?
A: Yes. Use controlled test patterns and native mapping to determine whether the symptom is fixed to the panel or follows the content, scaler, cable, player, or interface path.
Q:
A: A good pixel-defect specification replaces the sentence "no dead pixels" with a reproducible inspection contract. Define the categories, patterns, viewing condition, location and cluster logic, evidence, and supplier disposition before the shipment arrives. That is what turns a visual disagreement into a quality-control decision.
