What 2000 nits on a screen really means

You are standing in front of a wall of televisions. One tag shouts “2000 nits,” another boasts “HDR 1000,” a third promises “peak brightness 4000.” The salesperson assures you that more is always better. But what does the number actually mean — and is it worth paying for? The hero of this story is luminance: the physical quantity that decides how bright a screen looks to us. Along the way you will meet a unit with a surprisingly old pedigree, learn where the number π comes from in light measurement, and figure out how to read those advertised figures so they can't fool you.

Four concepts, one family

Photometry — the science of measuring light the way the human eye sees it — rests on four concepts. They are easiest to untangle using a candle.

The candela (cd) is luminous intensity: how much light a source sends in a given direction. An ordinary candle has a brightness of roughly one candela — which is where the name comes from (Latin candela = candle). It is one of the seven base units of the SI system. Since 2019 its definition has sounded rather esoteric: the candela is set so that the luminous efficacy of monochromatic radiation at a frequency of 540×10¹² Hz — yellow-green light with a wavelength of about 555 nm, to which the eye is most sensitive — is exactly 683 lumens per watt. In practice: a source emitting 1/683 of a watt of that green light per steradian in a given direction has a brightness of one candela.

The lumen (lm) is luminous flux — the total “amount” of light flowing out of a source in all directions. Lumens are what you find on a light-bulb box.

The lux (lx) is illuminance — how much light falls onto a surface (one lumen per square meter). This is the quantity a photographer's light meter and a phone's brightness sensor measure.

The candela per square meter (cd/m²) — and here we reach the hero — is luminance: the brightness of a surface that emits or reflects light. The distinction worth remembering forever shows up when you set it beside lux: lux describes light arriving at a surface, while cd/m² describes light leaving a surface toward your eyes. That is why luminance, and none of the other quantities, is responsible for the impression of “how bright this screen is.” The eye does not perceive “flux” or “illuminance” — it perceives the brightness of the surface it looks at. A screen, a sheet of paper, the sky out the window — all are surfaces of a certain luminance.

Why luminance and not lumens? Because two bulbs with the same flux but different size will produce a different impression of brightness — the smaller one will be “sharper,” more dazzling. Luminance accounts for the area the light flows out of, and so it best captures what we actually see.

The “nit” — the secret name of the candela per square meter

And now the promised surprise. Display makers almost never write “cd/m².” They write “nits.” And although it sounds like fashionable digital jargon dreamed up by 4K-TV marketers, the truth is quite different:

A nit is simply another name for the very same unit. 1 nit = 1 cd/m². Exactly, with no conversion factor.

The name comes from the Latin nitere — “to shine” (or the related nitor, “brilliance”). And it is anything but new. Merriam-Webster dates the first documented use of “nit” in this sense to 1953, and according to Russ Rowlett's respected Dictionary of Units of Measurement the name was approved by the International Commission on Illumination (CIE) as early as around 1947 — about when the 9th General Conference on Weights and Measures (1948) was formally establishing the candela itself. In other words: the “nit” is older than color television, let alone LCD or OLED. It is not a novelty, but a handy short label the display industry pulled out of the drawer and popularized. (Popular blogs and retailer pages sometimes cite the 1930s, but that date has no backing in any dictionary or standards document and should be treated as uncertain.)

There is an important caveat, though: the nit is not an official SI unit. It is a tolerated but “non-system” — and formally deprecated — name for cd/m². A scientist writes “cd/m²,” a display engineer writes “nit.” They mean the same thing.

That dispatches myth number one: that a nit is some new, digital unit. It isn't. It is the candela per square meter in disguise — a unit that has existed for decades.

Where does π come from in units of light?

Now the most interesting part technically. Alongside cd/m² and the nit, the world of luminance is home to a whole family of oddly defined units in which a factor of 1/π stubbornly appears:

apostilb = 1/π cd/m² ≈ 0.3183 cd/m²
foot-lambert (fL) = 1/π cd/ft² ≈ 3.426 cd/m²
lambert = 1/π cd/cm²
stilb (sb) = 1 cd/cm² = 10,000 cd/m² (a CGS unit)

Notice that the stilb contains no π — it is “honest”: one candela per square centimeter, full stop. But the apostilb, lambert and foot-lambert have that π built into their definition. Where did it come from?

The answer lies in Lambert's law and in geometry. Picture a perfectly matte, diffusing surface — a so-called Lambertian surface (fresh snow, matte paper, milky glass come close). Such a surface has a wonderful property: it looks equally bright from every angle. Look at it head-on or from the side — the luminance is the same. This happens because at an angle you see less light (it falls off as the cosine of the angle), but you also see a correspondingly smaller area (again as the cosine) — and the two effects cancel out exactly.

The problem appears when we want to sum all the light radiated by such a surface into the upper hemisphere. We then have to integrate that cosine distribution over all directions, and the result is not 2π (the full solid angle of a hemisphere) but exactly π. The consequence is surprising: a Lambertian surface that gives off 1 lumen per square meter into the whole half-space has a luminance of 1/π cd/m² — not 1 cd/m², as naive intuition would suggest.

The creators of the old “Lambertian” units treated this as a convenience. Since a surface illuminated at 1 lux (with full reflection) has a luminance of 1/π cd/m² anyway, let's define a unit so that the π vanishes from the arithmetic. Hence the apostilb, lambert and foot-lambert: they have π “hidden” in their definition, so that in practical lighting calculations it cancels against the π from the integration — in these units the luminance of an ideal diffuser is numerically equal to the illuminance falling on it. The SI system took the road of more honest physics and does not hide π in the definition of cd/m². That is why the conversions look so odd: 1 foot-lambert is 3.426 cd/m², not some round number.

The scale of luminance: from the night sky to the surface of the Sun

Luminance spans a staggering range — from fractions of a cd/m² to billions. Here is an ordered scale (these are orders of magnitude, strongly dependent on conditions):

Object / situation	Luminance (cd/m² = nits)	Roughly in fL
Night sky (zenith, far from cities)	~0.0002	negligible
Digital cinema screen (white, DCI/SMPTE)	48	~14
Screen calibrated for color work (SDR)	80–120	23–35
Typical monitor or living-room TV (SDR)	200–300	58–88
HDR peak of a modern TV (small window)	1500–3600	440–1050
Full Moon (the disc)	~2500	~730
Sheet of white paper in sunlight	~25,000	~7300
Surface of the Sun (at zenith)	~1.6×10⁹	~4.7×10⁸

For completeness: the stilb is 10,000 cd/m² (a sheet of paper in the sun is about 2.5 stilbs), while the apostilb is a fraction of a cd/m², so it suits dim surfaces better.

A few numbers worth remembering. The human eye registers luminances from about one-millionth of a cd/m² up to about one million cd/m² — above that, retinal damage looms. The Sun, with a luminance on the order of 10⁹, is roughly a thousand times brighter than that limit; that is why staring at it so easily ruins your eyesight. Notice, too, how low on this scale a “correctly” set screen sits: a monitor for professional color work is calibrated to just 80–120 cd/m², and digital cinema shows a white image at a luminance of a mere 48 cd/m². Brightness is not everything — context and contrast matter.

Why cinema is calibrated in foot-lamberts

Speaking of cinema — here is a tasty detail. Digital cinema worldwide is calibrated to 14 foot-lamberts (≈ 48 cd/m²) of white on the screen, measured “from the projector with no film in the gate” (an SMPTE standard, adopted by the Digital Cinema Initiatives spec with a ±3 fL tolerance). Historically, SMPTE 196M spoke of 16 fL with an empty gate, which — after accounting for the base density of the film stock — gave those 14 fL on the screen with film in place.

That is surprisingly little — several times less than an ordinary living-room TV. But cinema operates in perfect darkness, where the eye adapts to low levels and what counts most is contrast, not raw brightness. And why foot-lamberts rather than nits? Because it is a unit of a “reflecting screen”: a cinema image does not glow on its own — it reflects the projector's light. The foot-lambert conveniently ties together the projector's flux, the screen size and its gain — more on which in a moment.

“More lumens = a brighter image”? Not in a projector

This brings us to myth number two: “more lumens = a brighter image.” Lumens — or, more precisely, ANSI lumens — describe the total flux a projector throws onto the screen. But the perceived brightness of a surface is described by luminance, and that depends on more than the projector.

The same projector casting an image onto a small screen gives a brighter image than onto a large one — because the same flux is spread over a larger area. Screen gain also comes into play: a screen with gain 1.3 reflects 30% more light toward the viewer than a neutral one. An example from industry calculations: a 2000-lumen projector on a screen of about 13 m² yields roughly 14 fL; swap the fabric for one with gain 1.3 and the result rises to about 18 fL — with no change to the projector. That is why the foot-lambert, not the lumen alone, rules the world of projection and cinema: it tells you how bright the image you'll actually see will really be.

“A 2000-nit TV always hits 2000 nits”? Absolutely not

And finally myth number three, the most important when buying: that a “2000-nit” TV puts out 2000 nits whenever it likes. It does not.

Makers almost always quote peak luminance on a small window — a bright rectangle covering 1, 2, 5 or 10% of the screen against a black background. That is the figure for small highlights: the flash of an explosion, the sun glinting off glass, a spark. As the bright area grows, luminance drops dramatically. The culprit is ABL (automatic brightness limiter) — a circuit protecting the panel from overheating and exceeding its power budget.

The numbers speak for themselves. LG's flagship OLED G4: as Digital Trends measured, peak brightness was about 1500 nits with windows up to 10% of the screen, then fell to about 235 nits at full screen — roughly a sixfold drop. That is typical OLED behavior. Smartphones are similar: the Samsung Galaxy S24 Ultra advertises a 2600-nit peak (DXOMARK confirmed a measured 2572 nits in sunlight, “close to the advertised 2600”), but in a full-white-screen test PhoneArena measured about 1280 nits, and the iPhone 15 Pro Max about 1092. Records like “4500 nits,” meanwhile, can be pure marketing: TechRadar's reviewers could not squeeze even 1200 nits out of the OnePlus 13 in the lab “under any conditions,” despite its claimed 4500-nit peak.

Here a fundamental difference between technologies emerges. OLED and QD-OLED shine with pixels that emit their own light — giving perfect black and practically infinite contrast, but their full-screen brightness is limited by power and thermals (hence the aggressive ABL). Mini-LED is a backlit LCD with hundreds or thousands of dimming zones — it can sustain a far higher full-screen brightness (flagships reach peaks on the order of 3000–3600 nits, and 700–800 nits full-screen), at the cost of poorer black and a risk of “halo” around bright objects. That is why, when buying, it is worth looking not at a single peak number but at brightness across different window sizes — data published by RTINGS, FlatpanelsHD and Tom's Guide.

Real HDR versus marketing HDR

It is worth demystifying HDR here. The PQ standard (SMPTE ST 2084), on which HDR10 and Dolby Vision are based, maps signal values directly to absolute luminance levels — up to a theoretical 10,000 cd/m². In practice, films are usually mastered on reference monitors with a peak of 1000–4000 nits (Dolby requires at least 1000 nits of peak and a 200,000:1 contrast from such a monitor; the famous Dolby Pulsar reaches 4000). A screen that merely brightens an ordinary SDR signal to 1500 nits is not an HDR screen — real HDR requires support for the PQ or HLG curve, a wider color gamut and content prepared for that range. The VESA DisplayHDR certifications (400, 600, 1000, 1400 for LCD; True Black for OLED) are meant to bring order to this — though, as reviewers themselves note, even they do not guarantee identical results.

What it's all for: sunlight readability and white for calibration

Luminance has two very practical faces.

The first is readability in bright surroundings. It is not the screen's luminance alone that decides whether you'll see the image on the beach, but its ratio to the light reflected off the screen's surface. In full sun (illuminance on the order of 100,000 lux), a screen with 5% reflectance bounces about 5000 cd/m² of the surroundings alone back into your eyes — to “beat” that you'd need an unrealistic tens of thousands of nits. That is why matte, anti-reflective coatings matter as much as raw brightness: a good anti-reflective coating can do more for readability than another few hundred nits. Contrast in ambient light beats raw brightness.

The second face is calibration. For professional color work — photo editing, video grading — the screen is set to a surprisingly low, tightly specified white luminance. The ISO 12646 standard for “soft proofing” gives 80–120 cd/m², the sRGB spec aims for 80 cd/m², and reference SDR and television mastering (Rec. 709 / BT.1886) is 100 cd/m². All these values are far below a TV's default “showroom” brightness — because here what matters is fidelity and repeatability, not the “wow” effect.

What to actually check when buying a bright screen

Let's pull this into a practical conclusion. “2000 nits” on the box is not a lie — but it is a half-truth, because it refers to a small window. Instead of latching onto one number:

Check full-screen brightness, not just the peak on a 2–10% window. For sports, bright scenes and daytime work, that is what counts. A rule of thumb: good full-screen HDR brightness is around 600 nits or more in a bright room.
Match the technology to the room: Mini-LED with a good anti-reflective coating for very bright living rooms; OLED or QD-OLED, with their perfect black, for darkened home-cinema rooms.
Coating and reflections often matter more than extra nits — matte versus mirror.
Don't pay for “HDR” without substance: look for PQ/HLG support and a meaningful certificate (DisplayHDR 600 is a reasonable threshold for real HDR, 1000+ is the higher tier), not just a number.
Trust independent reviewers' measurements (RTINGS, FlatpanelsHD, Notebookcheck, DXOMARK), who report luminance across different window sizes, not just the marketing peak.

Because luminance is not a magic number on a label — it is the physics of light leaving the screen toward your eyes. The next time you stand before a wall of televisions, you'll know that a “nit” is the venerable candela per square meter in a new suit, that π hides in foot-lamberts thanks to Lambert's law, and that “2000 nits” must be read with the footnote “on a small window.”

Nits, candelas and the mysterious π — what “2000 nits” really means