How Many Nits Does an Outdoor LED Facade Need? A Specification Guide
Nits are the most quoted number in a facade quotation and the least understood. Here is how much brightness a surface actually needs, why 8,000 nits can look worse than 6,000, and how to write a brightness requirement a supplier cannot game.
Somewhere in every facade tender there is a line that reads “brightness: 6,000 nits or above”, written because it sounded safe. It is a number without a viewing condition, a reflecting surface or a dimming curve attached to it, and it protects nobody. A facade can be specified at 6,000 nits and still look washed out at noon, and a facade specified at 8,000 nits can be a nuisance complaint at nine in the evening.
This guide works through the actual physics in the order a project hits it: what a nit measures, why contrast rather than brightness decides legibility, how orientation changes the answer, and what to write in the specification so that the number you sign for is the number you get on the wall.
1. The short answer, by condition
| Condition | Sustained brightness | Why |
|---|---|---|
| Shaded elevation, arcade, deep recess | 2,500–4,000 nits | Little direct sun; more brightness adds heat, not legibility |
| North-facing wall (northern hemisphere) | 3,500–5,000 nits | Mainly ambient sky light, no direct solar load |
| Standard urban elevation, partial sun | 5,000–6,500 nits | The working range for most commercial facades |
| South or west facing, low latitude | 7,000–10,000 nits | Direct low-angle sun for hours, high reflected load |
| Glass curtain wall behind the screen | +20–40% on the above | The glass behind and beside the modules adds reflection the viewer must look through |
| Night operation | 100–400 nits | Regulation and comfort; running at day level at night is the most common complaint source |
Table 1. Planning ranges for sustained white brightness. These are starting points for a site assessment, not a substitute for measuring the actual elevation.
Notice what is missing from that table: the word “peak”. Almost every datasheet figure is a peak, measured on a small area, at a defined current. What your viewer sees is sustained full-field brightness at the drive current the product actually runs. On a facade, the difference between the two can be 25–40%.
2. What a nit is, and what it is not
A nit is one candela per square metre (cd/m²). It measures luminance: light leaving a surface in a direction, per unit of area. It is the right unit for a display because the viewer sees a surface, not a beam.
Three units get confused constantly in procurement documents:
- Lumen measures total light output of a source. It describes a lamp, not a screen. A facade datasheet quoting lumens is describing its LEDs, not its appearance.
- Lux measures light falling onto a surface. Sunlight on a facade might be 100,000 lux; that tells you how hard the environment is fighting the screen, not how bright the screen is.
- Nit measures light leaving the surface toward the eye. This is the one that predicts whether content is readable.
The relationship that matters is not brightness alone but the ratio between emitted and reflected light at the eye. That is the next section, and it is where most specifications fail.
3. Ambient contrast ratio: why 8,000 nits can lose to 6,000
Daylight does not just compete with a screen; it reflects off it. Sunlight hitting the module face, the mask between pixels, the encapsulation and any glass in front bounces back toward the viewer and adds a veil of light over the image. Legibility depends on how much of the emitted light survives that veil.
The engineering expression is the ambient contrast ratio:
ACR = (Ldisplay + Lreflected) ÷ Lreflected
Where Ldisplay is the emitted luminance and Lreflected is the luminance the surface throws back at the viewer under the ambient light of that site.
Put numbers in and the surprise appears. Take a bright elevation where reflected luminance reaches 1,200 nits — realistic on a pale mask under direct sun:
| Screen | Reflected load | ACR | What the viewer sees |
|---|---|---|---|
| 6,000 nits, deep black mask, matt finish | 700 nits | 9.6 : 1 | Clean image, strong blacks |
| 6,000 nits, glossy mask, untreated front | 1,200 nits | 6.0 : 1 | Readable but flat; colours shift |
| 8,000 nits, glossy mask, untreated front | 1,600 nits | 6.0 : 1 | Same contrast as the cheaper screen, more power and heat |
| 8,000 nits, anti-reflective front, deep mask | 750 nits | 11.7 : 1 | Best legibility of the four, highest cost |
Table 2. Illustrative ACR calculation. The point is not the exact numbers but the direction: surface treatment can move contrast as much as a 33% brightness increase, and it costs less in power and heat.
Two conclusions follow, and both are commercially useful.
First, surface treatment is a lever. A deep mask between pixels, a matt encapsulation and an anti-reflective front layer all reduce the reflected term. Going from 8% to 2% effective reflection is worth roughly the same as doubling emitted brightness in perceived contrast — without doubling the electricity bill or the junction temperature.
Second, the headline number alone cannot be compared across suppliers. Two quotations at 6,000 nits can differ by 50% in daytime legibility depending on mask design, encapsulation and front glass. If a supplier will not state the mask and the surface treatment, the nit figure is marketing.
4. Where the reflection comes from
Reflected luminance is not one number; it is the sum of several surfaces, and each can be engineered.
| Surface | Contribution | What can be done |
|---|---|---|
| Pixel mask (the black area between LEDs) | Dominant at coarse pitch, because there is more mask than emitter | Deep black matt mask; avoid pale or glossy resin. On a P50 mesh this single choice changes the daytime look more than 1,000 nits |
| Encapsulation over the LED | Moderate; glossy potting acts as a small mirror per pixel | Matt potting, or a textured surface that scatters rather than reflects |
| Front glass or polycarbonate, if used | High; untreated glass reflects 4–8% straight back | Anti-reflective coating or optical bonding; avoid a secondary glazing layer in front of the screen |
| Building glass behind and beside a transparent screen | Site-specific and often underestimated | Nothing on the screen side; this is a viewing-position problem, solved by content and by accepting a lower contrast at grazing angles |
| Adjacent pale surfaces: paving, render, water | Can double the effective ambient load at eye level | Not controllable; must be modelled in the brightness decision rather than discovered at commissioning |
Table 3. The reflection budget. On coarse-pitch facade products the mask is usually the largest single term, which is why pitch and legibility are linked.
5. Orientation and latitude: the sun is not symmetrical
Two elevations on the same building can need different specifications. This is obvious to a lighting designer and routinely absent from procurement documents.
In the northern hemisphere, a south-facing elevation receives direct sun for most of the day, and at a low angle in winter when the sun is close to the horizon and the viewer’s eye line. A west-facing elevation takes the afternoon sun, which is often the worst case because it coincides with the time people are on the street and the light is low and intense. A north-facing elevation never sees direct sun and is governed by sky luminance.
The practical implications:
- A single brightness figure across a multi-orientation building is a compromise. Either the south face is marginal at noon or the north face is over-specified and over-heated. Splitting the specification by orientation is legitimate and usually cheaper overall.
- Low-latitude sites need more headroom. Near the equator the sun is high, but the ambient load is high all day and the diffuse component is large. Planning at 7,000–8,000 nits for a fully exposed elevation is not extravagant.
- Reflected sun from a neighbouring glass tower is a real case. A curtain wall opposite can throw a specular reflection onto the elevation for an hour a day. It is worth asking during the site survey, because no amount of brightness wins against a direct specular hit — the answer is content design and dimming, not more nits.
6. Peak versus sustained: the specification trap
A datasheet typically states brightness at a reference current, sometimes at a reduced duty cycle, occasionally on a white field at a colour temperature that flatters the measurement. Three variants hide behind one number:
- Peak small-area brightness. Measured on a patch or a single module. Higher than what the whole screen can hold.
- Sustained full-field brightness. The whole array at white, at the drive current the system actually runs. This is what a viewer sees.
- Brightness after thermal derating. What the screen still does at 45°C ambient after an hour, when the driver has folded back to protect itself.
On a facade, the third number is the one that matters in August. We specify sustained full-field figures at a stated ambient temperature, and we expect a supplier who cannot do the same to be hiding a derating curve.
The way to close the gap is contractual, not technical: require the brightness figure to be stated as sustained full-field at a named ambient temperature, verified on the production sample with a calibrated meter, and repeated in the pre-shipment inspection.
7. Brightness costs power, and power costs money every year
LED output scales roughly with drive current, and so does power. Doubling brightness does not double the bill — efficiency falls at high current — but it is close to linear in the range facades operate.
Planning figures for outdoor LED at typical content (not full white):
| Sustained brightness | Average power | 500 m² at 8 h/night | Annual cost at $0.15/kWh |
|---|---|---|---|
| 3,000 nits | ~110 W/m² | 440 kWh/night | ~$24,100 |
| 5,000 nits | ~170 W/m² | 680 kWh/night | ~$37,200 |
| 7,000 nits | ~230 W/m² | 920 kWh/night | ~$50,400 |
| 10,000 nits | ~310 W/m² | 1,240 kWh/night | ~$67,900 |
Table 4. Illustrative operating cost for a 500 m² elevation. The delta between 5,000 and 10,000 nits is roughly $30,000 a year — which is why over-specifying brightness is a permanent cost, not a one-off.
The corollary is that dimming is a revenue decision as well as a compliance one. A screen that runs at 5,000 nits at noon and 300 nits at midnight uses a fraction of the energy of one that runs flat out from dusk. Automatic dimming with a light sensor, not just a time schedule, is standard on any facade we build, and its absence should be a red flag in a quotation.
8. Brightness, heat and how long the wall lasts
Running LEDs harder shortens their life. A 100,000-hour L70 rating is normally measured at a defined current, often 50% of maximum. Drive continuously at 100% and the practical life can halve. On a facade, where replacement means access equipment, that is an expensive way to save on modules.
Three design habits follow:
- Specify headroom, then run below it. A 7,000-nit screen run at 5,000 nits for most of its life will outlast a 5,000-nit screen run at maximum.
- Treat thermal design as part of the brightness specification. An over-driven module in a sealed cabinet derates itself; the manufacturer’s answer is often a bigger heatsink or active cooling, both of which cost money that should be in the original comparison.
- Ask for the derating curve. A supplier who can show brightness against ambient temperature and against drive current is telling you they measure their product. One who cannot is quoting a lab condition.
9. Night is the other half of the specification
Every facade has two operating modes and most specifications only describe one. At night a 6,000-nit elevation is not merely wasteful; it is a nuisance, a traffic-safety concern and, in many jurisdictions, an enforcement risk.
Common regulatory expectations, and sensible practice regardless:
| Condition | Typical requirement | Practical implementation |
|---|---|---|
| Curfew | Reduced output or static content after 22:00–23:00 | Scheduled scene in the media server, with an override log |
| Luminance cap | Often expressed as a maximum night luminance for the surface | Dimming curve that targets a measured level, not a percentage of maximum |
| Response to ambient light | Output should follow sky luminance, not the clock alone | Light sensor on the elevation, calibrated at commissioning |
| Light spill toward residences | Avoid direct view of high-luminance surfaces from windows | Content masking, louvres, or a physical setback; not achievable by dimming alone |
| Content restrictions | No rapid flashing near roads; no high-contrast strobe | Content policy in the media server rules, checked at approval stage |
Table 5. Night-time obligations that are cheaper to design in than to retrofit after a complaint.
The design point is that a dimming curve is not one setting. A good facade has at least three: a commissioning profile, a night profile and an event profile. If the control system cannot store three, it will be run at one, and the one will be the day setting.
10. The special case: a screen on a glass facade
Glass changes the calculation twice, and both are usually missed.
Behind the screen. A transparent or mesh product seen against a lit interior has two competing images: the emitted content and whatever is happening behind it. Contrast falls not because the screen is dim but because the background is not black. This is why a transparent facade usually needs higher brightness than a solid one in the same location, and why the content has to be designed for a background that moves.
Beside the screen. Glass reflects. A neighbouring curtain wall throws light onto the elevation, and a glossy front surface reflects the sky. Neither is solved by nits; both are solved by surface treatment and viewing position.
Practical guidance: on glass facades, model the brightness against the actual interior lighting and the reflection from adjacent glazing, and accept that daytime contrast will be lower than a solid elevation. Content with large flat fields and strong colour separation reads far better than fine photographic detail, which is a content decision with a brightness consequence.
11. How to write the requirement
A brightness clause a supplier cannot game looks like this. Ten lines, all verifiable.
- Sustained full-field white brightness at a named ambient temperature (for example 40°C), stated in nits, not peak.
- Measured on the production sample with a named instrument type, at a named distance, with the result in the pre-shipment report.
- Mask specification: colour, finish and reflectance class.
- Encapsulation finish: matt or glossy, with the material named.
- Front surface: whether any glass or coating is present, and its reflection figure.
- Dimming range: the minimum and maximum, as a percentage of full output, and whether it is linear or perceptual.
- Dimming control: light sensor plus schedule, with the sensor model named.
- Derating curve: brightness against ambient temperature and against drive current, supplied as data.
- Power at the specified brightness: watts per square metre at content average, and maximum.
- Uniformity: brightness variation across the module and module-to-module, as a percentage.
If a supplier answers nine of these and avoids the derating curve, that is the one to press on. It is the line that separates a measured product from a datasheet.
12. Five mistakes we see repeatedly
1. Specifying a single number for the whole building. South and north elevations do not need the same brightness; one will be marginal and the other over-heated.
2. Comparing peak figures across suppliers. Peak small-area numbers are not comparable and are not what the viewer sees. Ask for sustained full-field.
3. Buying brightness instead of contrast. A better mask and a matt encapsulation often beat another 1,500 nits on both legibility and running cost.
4. Ignoring the night profile. The most common post-installation complaint is glare, and it is almost always a dimming design failure rather than a hardware one.
5. Forgetting that brightness is a lifetime cost. The difference between 5,000 and 10,000 nits is tens of thousands of dollars a year on a large elevation, plus heat and shortened module life.
13. Frequently asked questions
How many nits does an outdoor LED screen need in daylight?
For a standard urban elevation with partial sun, plan 5,000–6,500 nits sustained. A south or west facing elevation at low latitude needs 7,000–10,000. A shaded or north-facing elevation is usually fine at 3,000–5,000. These are sustained full-field figures, not peak small-area numbers.
Is 5,000 nits enough for a facade in direct sunlight?
It depends on the reflected load. In direct sun with a pale or glossy mask, 5,000 nits can look washed out; the same 5,000 nits with a deep mask and matt encapsulation can read cleanly. Contrast, not brightness alone, decides legibility, which is why the surface specification belongs in the same clause as the nit figure.
Is more nits always better?
No. Beyond the point where contrast is adequate, extra brightness adds power, heat, shorter module life and night-time glare. On a facade, the design target is the lowest brightness that gives a clean image at the worst viewing condition — not the highest number in the brochure.
What is the difference between peak and sustained brightness?
Peak is measured on a small area, sometimes at reduced duty cycle or at a chosen colour temperature. Sustained full-field is the whole array at white, at the drive current the system runs. On outdoor products the gap is commonly 25–40%, and only the sustained figure predicts what a viewer sees.
What is ambient contrast ratio and why does it matter?
ACR is the ratio of emitted plus reflected light to reflected light at the viewer: (Ldisplay + Lreflected) ÷ Lreflected. It predicts legibility better than brightness, because a screen in a highly reflective environment loses contrast even if it is bright. Raising ACR by improving the mask can be cheaper than raising nits.
How bright should a facade be at night?
Typically 100–400 nits, with a curfew reducing output or freezing content after about 22:00–23:00. Many authorities cap night luminance for illuminated surfaces, and some require advertisement consent for any illuminated elevation regardless of content. Automatic dimming with a light sensor is standard practice.
Does brightness affect how long the LEDs last?
Yes. Running at maximum drive current shortens life; the common L70 rating is measured at a defined current, often 50% of maximum. Specifying headroom and running below it is cheaper over a facade’s life than replacing modules that need access equipment to reach.
How much power does a bright facade use?
At typical content, roughly 110 W/m² at 3,000 nits rising to about 310 W/m² at 10,000 nits. For a 500 m² elevation at eight hours a night, the annual difference between 5,000 and 10,000 nits is around $30,000 at $0.15/kWh.
Do transparent screens need more brightness than solid ones?
Usually yes, because the surface is seen against a background that is not black, so contrast falls for reasons unrelated to the screen. A glass facade also adds reflection from adjacent glazing. Content with large flat fields and strong colour separation compensates better than fine photographic detail.
Can I just turn the brightness up after installation?
Only within the headroom designed in. A module specified at 5,000 nits cannot be pushed to 8,000 without over-driving it, which increases heat and shortens life. If there is any chance the site will need more, specify the higher figure at the start and dim down — not the other way round.
How do I verify the brightness I paid for?
Require it in three places: the datasheet as sustained full-field at a named ambient temperature, the golden sample measured with a calibrated meter, and the pre-shipment inspection report with the same method. A specification that only appears in the brochure is a claim, not a requirement.
Does pixel pitch change the brightness requirement?
It changes the perceived result. Coarse pitch has more mask area, so it reflects more and has less emissive area; a P50 mesh often needs a higher nit figure than a P16 panel to deliver the same daytime impression. This is why brightness should be specified together with pitch and mask, not on its own.
Need a brightness figure for a specific elevation?
Send the elevation orientation, the city, the viewing distance and a representative frame of the intended content. We will return a sustained brightness target, the mask and surface treatment that support it, and the power and dimming profile that follow from it.
Request a brightness assessment Outdoor architectural lighting
Project enquiry
Send your drawings. Get a specification you can defend.
Elevation DWG, facade area, closest viewing distance and site country — we reply within 48 hours with a fixture matrix, power and DMX topology, and a quotation. No price-only quotes.
5,000+ projects delivered · 40+ countries · ISO 9001 · CE / UL / RoHS


