If you have ever tried to figure out whether a digital sign vendor’s brightness spec actually satisfies a municipal code, you have run into a problem that has nothing to do with the sign itself: the number on the spec sheet and the number in the ordinance are frequently measured in two entirely different systems. We pulled every digital-sign brightness rule in our corpus to see how common each measurement system actually is, and the answer explains a lot of confused permit conversations.
Most codes don’t quantify brightness at all
Our corpus contains 1,131 rows tagged as digital-sign brightness rules. Of those, only 188 carry any machine-parseable unit at all. The other 943 — 83.4%, five out of every six rows — describe brightness in purely qualitative language: phrases like “shall not create a glare” or “shall not be unreasonably bright,” with no number attached at all. This is not an extraction failure on our end; it is an accurate reflection of how most sign codes are actually written. Brightness is one of the attributes municipalities are most likely to leave to subjective enforcement rather than a hard numeric standard.
Two incompatible systems, both in active use
The 188 rows that do carry a unit split into two measurement families that are not directly comparable to each other, plus a small residue that belongs to neither:
| Unit family | Rows | Share of unit-bearing rows |
|---|---|---|
| Nits family (nits, cd/m², candelas per square meter) | 119 | 63.3% |
| Footcandle family (footcandles, foot-candles, ft-candles) | 62 | 33.0% |
| Neither (percent of maximum, lumens, mis-parsed units) | 7 | 3.7% |
- Nits (candelas per square meter) — a measure of the sign’s own absolute light output, independent of its surroundings. This is the unit digital sign manufacturers use on spec sheets, and it is the same unit used for TV and phone screen brightness.
- Footcandles above ambient — a measure of how much additional light the sign adds relative to the ambient light level already present at the sign’s location. This is a relative measurement: the same physical sign could be compliant in a brightly lit commercial corridor and non-compliant in a dark rural setting, because the ambient baseline is different.
These are not two ways of expressing the same quantity — they measure genuinely different things. A brightness value in nits tells you nothing about whether a sign clears a footcandle-over-ambient limit, and vice versa, without a light-meter reading taken at the specific installation site under its actual ambient conditions.
A detail worth noting before the distributions: those 188 rows are written in 33 different spellings of the unit — “nits,” “NITS,” “NITs,” “cd/m2,” “candelas per sq meter,” “footcandle,” “foot candles,” “ft-candles,” and two dozen more. Counting only the exact string “nits” gets you 93 rows and loses a fifth of the family; an earlier version of this article made exactly that mistake and published a split that did not add up. It is the same vocabulary problem we documented in forty names for the same sign, showing up one level further down, in the units rather than the sign types.
The nits distribution
Of the 119 nits-family rows, 118 carry a usable number:
| Percentile | Nits |
|---|---|
| Minimum | 100 |
| 10th percentile | 250 |
| 25th percentile | 500 |
| Median (50th) | 5,000 |
| 75th percentile | 5,000 |
| 90th percentile | 5,000 |
| Maximum | 12,000 |
The jump between the 25th percentile (500 nits) and the median (5,000 nits) is enormous, and the fact that the median, 75th, and 90th percentile all land on exactly the same value is not a coincidence: 58 of the 118 numeric rows — 49.2% — sit at exactly 5,000 nits. That is a de facto industry cap that a large share of codes have adopted outright rather than an organically distributed value. If you are speccing a digital display for a code that caps brightness in nits, 5,000 is the number to check against first. The low end of the range is usually measuring something different: of the 44 nits-family values below 1,000, 30 explicitly reference night-time, dusk, or after-sunset operation in the rule text. Read a 250-nit limit as a curfew provision, not as an unusually strict daytime cap.
The footcandle distribution
The 62 footcandle-family rows (61 with a usable number) measure brightness as an increase above ambient light:
| Percentile | Footcandles above ambient |
|---|---|
| Minimum | 0.1 |
| 10th percentile | 0.2 |
| 25th percentile | 0.3 |
| Median (50th) | 0.3 |
| 75th percentile | 0.3 |
| 90th percentile | 0.5 |
| Maximum | 70 |
The clustering here is even tighter: 42 of the 61 numeric rows — 68.9% — sit at exactly 0.3 footcandles above ambient. The maximum of 70 is not a real regulatory outlier so much as a warning about the data: a value that far above the 90th percentile of 0.5 is much more likely to be a mis-parse of the ordinance text than a code that permits a sign 140 times brighter than its neighbours’. We keep it in the distribution rather than quietly dropping it, for the reason described in what a sign code doesn’t tell you — the extremes of any extracted distribution are where you should be most suspicious, and hiding them makes the data look better than it is.
Why this is a genuine interpretive problem, not a formatting quirk
The practical consequence of these two systems coexisting is that you cannot treat a brightness number as portable across codes without first checking which system it is expressed in. A digital sign rated at 5,000 nits by its manufacturer is describing the sign’s own output in isolation. A code that instead caps brightness at 0.3 footcandles above ambient is asking a completely different question: how much brighter is this sign than the environment around it, measured on site, at night, with a light meter. Satisfying the first does not tell you anything about the second. A technician quoting a job in a footcandle-regulated jurisdiction who only checks the manufacturer’s nits rating has not actually verified compliance — they have verified a different, unrelated number.
And because 83.4% of brightness rules in our corpus don’t specify a unit at all, the most common real-world situation is neither of these systems: it is a qualitative standard (“no glare,” “not unreasonably bright”) that leaves the actual compliance determination to the zoning administrator or code enforcement officer at inspection time. In that situation, the safest approach is not to hit a specific number at all, but to design toward the visibly more conservative end of the range and be prepared to adjust based on inspector feedback. The other electronic-message-centre rules — dwell time, transition effects, curfews — are quantified far more often; we cover those in digital sign rules nationwide and digital sign dwell time.
How to check the brightness rule for your project
A report shows the digital sign brightness requirement for the applicable zone, including which measurement system the ordinance actually uses, cited to the ordinance section. If the code uses qualitative language rather than a number, the report shows that too, rather than forcing a number where the ordinance doesn’t provide one. Creating an account is free.