Quick verdict by scenario
The single biggest mistake in this comparison is assuming a dead pixel is a dead pixel. It is not. The same physical defect — one LED that never lights, one that never stops lighting — carries a very different acceptance consequence indoors and outdoors, because four things change at once: the pixel pitch you are buying, the number of LEDs in the field of view, how far people stand from the screen, and how you get access to the module to fix it.
- Close-viewing indoor screens (fine pitch, roughly P2.5 and below, audience within a few metres): treat dead pixels as a near-zero-tolerance image-quality issue and test at the worst-case viewing distance, not at the back of the room. A cluster of three adjacent black dots is visible; a scattered tiny percentage failure rate is what the eye actually notices.
- Outdoor screens (coarser pitch, audience tens of metres away, direct sun): the same absolute pixel count is far less visible from the footpath, but the defect that matters most is the stuck-on bright dot, which reads as a permanent white speck against a dark background at night — and outdoor cabinets make it harder to reach the module for repair.
- Mixed sites (stadium concourse, transport hub, retail entrance): write two separate acceptance clauses for the same physical screen if part of the audience is close and part is far, or you will be arguing about which number applies.
In short: indoor specs should be written around visibility at the closest realistic viewing distance; outdoor specs should be written around bright-dot counts, module-level replaceability and water ingress protection, because every pixel-level repair outdoors risks re-opening the seal.
Why the criteria are not interchangeable
A dead-pixel clause in a specification is not a single number. It is a bundle of five decisions, and each one changes with the indoor/outdoor context:
- What counts as a defect. Dark dots (blind, never lit), bright or stuck-on dots (always lit), and partial sub-pixel failures are usually counted separately. Failure to define which behaviour counts — under which test pattern — is the most common source of disputes.
- The denominator. Defects are expressed per module, per square metre, or as a ratio of LEDs. The same ratio produces wildly different absolute allowances on a 0.5 m² indoor cabinet at P1.5 and a 2 m² outdoor cabinet at P10.
- Clustering rules. Most buyers discover after delivery that a small percentage allowance legally permits a knot of failing pixels in the middle of the picture. Adjacency rules (no more than n defective pixels within a defined radius, and none at all near the screen centre) are what actually protect image quality.
- The test conditions. Solid-colour fields, a defined drive level and brightness setting, a dark-adapted assessor, and a fixed viewing distance and angle. Without these, two inspectors will always disagree.
- Remedy. Whether a failed pixel is repaired on site, the module is swapped, or the cabinet is returned, and who pays the rigging and access cost.
Only the third and fourth items are genuinely technical. The rest are commercial agreements about what the buyer will accept — which is why a straight copy-paste of an outdoor clause into an indoor specification almost always produces a number that is simultaneously too loose on visibility and too strict on administration.
Side-by-side: the same criteria, weighted differently
| Criterion | Indoor LED display (typical fine pitch) | Outdoor LED display (typical coarse pitch) |
|---|---|---|
| Typical pixel pitch | Roughly P0.9–P3, fine pitch for stages, control rooms, retail | Roughly P4–P20 depending on audience distance |
| Viewing distance rule of thumb | ~1 mm of pitch per 1 m of closest viewing distance (P1.5 ≈ 1.5 m) | Tens of metres; close inspection usually impossible and irrelevant |
| Dominant visibility risk | Single dark dots and clusters in mid-grey or flat-colour fields | Stuck-on bright dots at night; dust or moisture trapped behind the mask |
| Suggested tolerance posture | Very low, with hard adjacency rules and a zero-defect band in the central area | Slightly higher ratio defensible, but near-zero bright dots near eye level and in any camera-visible zone |
| Access for repair | Usually front or rear service; screens near human height, low access cost | Front service preferred; rigging, boom lift or scaffold needed, high access cost |
| Seal implications of repair | Minimal — indoor cabinets are not sealing-critical | Significant — every module swap risks the enclosure IP performance (IEC 60529) |
| Environment-driven defects | Dust, cleaning residue, thermal cycling in conditioned rooms | UV, rain, salt, heat, thermal cycling of large cabinets |
| Where clauses come from | Supplier module-level quality data; display-metrology terminology (VESA FPDM) | Supplier module data plus national LED screen standards that grade pixel out-of-control rates |
The table deliberately does not print specific pass/fail ratios. Published thresholds differ by standard, by screen class and by vendor, and quoting one number as if it were universal is exactly how acceptance disputes start. What transfers between projects is the structure of the clause, not the digit inside it.
Indoor: write the clause around the closest viewer
Indoor screens are the case where the industry rule of thumb about pixel pitch and viewing distance is most useful. If the front row of a conference room, a retail window or a broadcast studio sits roughly 1 m to 4 m from the glass, a P1.2 to P4 screen is the normal envelope, and the viewer can resolve individual pixels. That has three practical consequences for the dead-pixel clause.
1. Test at the closest distance, not the nominal one
A screen that looks clean from 8 m can look speckled from 2 m. Specify the test distance as the closest realistic view (front row, camera position, escalator landing) and state the pattern set: full red, green, blue and white fields, plus 50% grey — the last one is where clustering shows up most clearly. Ask the supplier to record which pattern revealed each defect.
2. Treat adjacency as the real limit
A ratio alone will pass a screen that has three adjacent dark dots beside a presenter's face. Add: no two defective pixels adjacent horizontally, vertically or diagonally; no more than a fixed small count within any 100 mm × 100 mm area; zero bright dots anywhere on the front face; and zero defects of any kind inside a defined central rectangle sized to the primary content area.
3. Separate the "as delivered" and "warranty" numbers
An indoor screen that is acceptable at handover can drift during the first year, particularly after thermal cycling. State an acceptance number at delivery and a drift allowance over the warranty period, with the trigger being either a count increase or the appearance of any bright dot — not a vague reference to "normal wear".
Indoor buyers also have an easier remedy conversation: modules are usually reachable from the floor or a modest platform, so the sensible default is module swap rather than field-level pixel rework, and the swap is cheap enough that insisting on near-zero tolerance is realistic.
Outdoor: write the clause around bright dots and re-sealing
Outdoors the arithmetic changes completely. A P10 cabinet has roughly a quarter of the pixel density of a P5 cabinet of the same area, so a percentage-based clause translates into a much larger absolute number of permitted defects on a big fascia — but those defects are usually invisible from 30 m away. What is not invisible is a bright dot burning through the picture at night, or a dead band along a module join that reads as a seam.
1. Bright dots deserve a stricter limit outdoors than indoors
Stuck-on pixels emit light against a dark scene and, on a high-brightness outdoor screen, are surrounded by high contrast. Buyers routinely accept a small dark-dot allowance outdoors and demand zero stuck-on pixels, particularly in the band between roughly head height and the top of the visible area from the nearest pedestrian route.
2. Module replaceability is part of the acceptance criteria
Outdoor screens are commonly specified with sealed cabinets rated to an IP level under IEC 60529 — IP65 (dust-tight, protected against water jets) is a common baseline, with higher second digits where wash-down or driving rain at an exposed site is expected. Whatever the exact rating, the practical question is: after a module swap, does the enclosure still meet it? Specify front-serviceable modules with factory-rated gaskets, state that the gasket is replaced during the swap, and require a post-repair visual check of the seal. Note that ratings beyond IPX6 are not cumulative — a cabinet rated for powerful jets is not automatically rated for the immersion tests — and salt-laden coastal air is outside the scope of the standard's fresh-water test conditions.
3. Expect environment-driven defects, and say how they are handled
Outdoor LED packages face UV exposure, wide temperature swings and, at coastal sites, corrosive atmospheres. Those are not pixel defects in the procurement sense, but they show up as failing pixels over time. A clause that only covers handover invites an argument at month fourteen. Define an inspection cadence (for example a documented walk-round from the nearest public viewpoint at a fixed brightness setting), a defect log format, and whether environmental degradation is warranty work or a maintenance item.
Trade-offs and exceptions
- Strict indoor tolerances cost money, and the cost is not linear. Pushing a fine-pitch screen to a near-zero defect rate pushes up module sorting and rework. Before writing a zero-bright-dot clause for a rental panel that lives in a flight case, ask whether the audience ever stands close enough to see it.
- Loose outdoor tolerances can be defensible — but only outside camera view. Any outdoor screen that is filmed, broadcast or photographed as part of the surroundings, such as an event façade or a stadium ribbon, needs indoor-style adjacency and central-area rules regardless of how far away the audience usually stands.
- Brightness is the hidden variable. Outdoor screens are driven far harder than indoor ones; high-luminance sources also fall under photobiological safety assessment (the IEC 62471 blue-light hazard framework uses a surface-luminance threshold around 10,000 cd/m² above which measured risk grouping is expected). That does not change the dead-pixel clause, but it is a reminder that pixel-level criteria sit inside a wider optical specification and should be tested at the drive level the screen will actually run at.
- Transparent mesh and flexible panels do not fit either column. Reduced LED density, cut-outs and curved substrates make clustering rules harder to apply; agree a pattern-by-pattern visual reference instead of a percentage.
- Cheap insurance: a photographed handover record of every test field, at the specified distance and brightness, is worth more than a tighter number in a contract you will struggle to enforce later.
How to decide next
- Measure the real viewing geometry — closest realistic viewer, camera positions, and the pedestrian route past the screen. This single input decides which column of the table above governs your specification.
- Decide the defect taxonomy first — dark dot, bright dot, adjacent group, partial sub-pixel — and only then pick numbers. Two documents: a one-page classification table and a test pattern list.
- Fix the test protocol — patterns, drive level, brightness setting, ambient conditions (dark-adapted for the visual check; note whether it is performed at night for outdoor), viewing distance and angle, and who signs.
- Write the remedy ladder — pixel rework, module swap, cabinet exchange — with who supplies the part, who pays for access, and a response time, because outdoor access is the slowest and most expensive step.
- Pair it with an environmental clause where relevant: the required IP code under IEC 60529, the seal check after module replacement, and a cleaning method that does not degrade the mask or drive defects.
- Ask for module-level rather than screen-level data from suppliers, so you can set the clause at the cabinet level and verify it per module at handover.
If you are comparing the two screen types for a whole project rather than an acceptance clause, work through brightness, pitch and mounting separately, and treat this dead-pixel clause as the last chapter — the one that determines whether the difference you paid for survives the first year of operation.
Sources
- IEC 60529 (EN 60529) — degrees of protection provided by enclosures (IP code): definitions of the first digit (dust-tight) and second digit (water jets, immersion), plus the note that ratings beyond IPX6 are not cumulative. Digit tables summarised at IP code.
- VESA Flat Panel Display Measurements Standard (FPDM), pixel-defect terminology — pixel or sub-pixel that fails to respond to addressing, bright (stuck-on) dot, dark dot; the standard defines measurement procedures rather than universal pass/fail limits.
- SJ/T 11141, General specification for LED display screens — grades the pixel out-of-control rate separately for indoor and outdoor screens, the closest thing to an authoritative split between the two classes.
- IEC 62471 / IEC TR 62778 — photobiological safety of lamps and lamp systems; surface-luminance threshold of around 10,000 cd/m² used to screen for retinal blue-light hazard, with risk-group classification.
- Manufacturer viewing-distance guidance (e.g. Christie Digital pixel-pitch viewing distance guidance) — the 1 mm pitch per 1 m minimum viewing distance rule of thumb used above.