Measuring Display Brightness: Methods, Instruments and Reporting

Brightness is the most quoted and least comparable display number. Two suppliers can report different figures for the same panel because they measured at different distances, after different warm-up times, at different points, or with instruments of different classes. A measurement method that travels with the number is what makes a brightness claim useful.

Part of the Optical, Backlight & Panel Technology guide — 30 articles on bonding, cover glass, brightness and panel choice.

What a nit actually is

One nit is one candela per square metre.

Luminance measures light emitted per unit area in a given direction, which is why a nit is a directional quantity and why the viewing angle belongs with any figure. A lumen, by contrast, measures total light output from a source in all directions, so lumens and nits cannot be converted without knowing the emitting area and the distribution. That distinction explains why a 3,000-lumen projector and a 300-nit display are not comparable numbers.

For a display, the useful figure is the luminance at the surface of the screen, measured from the direction the viewer will use.

Instruments: colorimeter, spectroradiometer, imaging colorimeter

Three classes, three levels of confidence.

A colorimeter is fast and inexpensive, with filters matched to a display’s spectrum; it is accurate on similar spectra and less so on narrow-band sources such as some LED backlights. A spectroradiometer measures the spectrum and is the reference instrument, slower and more expensive. An imaging colorimeter captures a whole screen at once, which makes uniformity and mura measurement practical rather than point-by-point. Instrument and procedure documentation is published by measurement vendors such as X-Rite, and detector options by photonics suppliers such as Hamamatsu.

Whichever is used, the instrument’s measurement aperture and its calibration status belong in the report.

Geometry: distance, angle and field of view

The geometry must match the application.

Luminance falls with viewing angle according to the panel’s characteristics, so a figure measured on axis does not describe what a driver sees from 30 degrees off. Measurement distance and the instrument’s field of view determine how much of the screen contributes, which matters near the bezel and on small panels. The practical rule is to state the angle, the distance and the field of view with every figure, and to measure at the angles the application actually uses.

Where an application is specified by contrast in ambient light, the geometry includes the illumination direction, because that is what produces the reflected component.

Warm-up and stability

A display measured cold is not the display the user sees.

LED output falls as the junction temperature rises, so a display measured immediately after switch-on reads higher than the same display after an hour at full brightness. The specification should state the warm-up condition, and measurements should be taken after the display has stabilised at the brightness setting used in service. The same applies to temperature: a measurement in a cold room and one in a warm enclosure will differ.

Recording luminance after a fixed settle time and at thermal equilibrium gives two figures that can be compared with another supplier’s.

Where to measure, and how many points

Centre-only figures hide the problems that annoy users.

A single centre measurement describes the brightest part of the panel; it does not describe the corners, where luminance is usually lower, or the uniformity that determines whether a large screen looks even. A standard grid – typically 3×3, 5×5 or 9×9 depending on the panel size – gives a set of values from which uniformity can be calculated and compared. CDTech’s notes on measuring brightness, contrast and uniformity describe the grid and the reporting practice.

Where a customer complaint concerns a region rather than the whole screen, an imaging measurement records it in a way that a point measurement cannot.

Contrast and the dark state

Contrast depends on the black level, which is where ambient light hurts most.

Contrast ratio is the white-field luminance divided by the black-field luminance, and both must be measured in the same conditions. In a lit room the black level rises because the front surface reflects ambient light, so the same panel reports a very different contrast figure in a dark room and in daylight. Colour and appearance definitions used for this work are published through the CIE, and haze measurement for the surface contribution follows defined methods such as ASTM D1003.

That is why a sunlight-readability specification is written as a contrast requirement in a stated ambient condition rather than as a brightness figure.

How many nits are needed for sunlight?

Derive the figure from the ambient light rather than looking it up.

The readable number depends on two things: the ambient illuminance in the plane of the screen and the reflectance of the front surface. A bonded anti-glare display with low reflectance can be readable at a few hundred nits in conditions where a glossy, unbonded display at over a thousand nits is not. The way to answer the question for a specific installation is to measure the ambient level, estimate the reflected luminance, and require emitted luminance high enough to hold the contrast the application needs.

Where a supplier quotes a nits figure without the conditions, ask for the measurement report, because the conditions are what make the number meaningful.

Reporting: what belongs with the number

Eight items make a brightness report reproducible.

• Instrument model, aperture and calibration date.
• Measurement distance, angle and field of view.
• Display brightness setting and warm-up time or thermal condition.
• Ambient illumination, and whether the display is in a dark room or lit environment.
• Test pattern: full white, full black, and the greys used.
• Measurement points and the grid used.
• Results per point, plus the calculated uniformity.
• Ambient temperature and, where relevant, the display’s own temperature.

A report with these items can be compared with another supplier’s, and it settles disputes without repeating the test.

Common measurement mistakes

Five errors that make figures incomparable.

Measuring cold and reporting as steady state; measuring on axis and applying the figure to an angled installation; using a colorimeter on a spectrum it was not calibrated for; placing the instrument at a distance where its field of view includes the bezel; and quoting luminance without stating whether the room was dark or lit. Each is easy to avoid and each is easy to detect in a supplier’s report once the eight reporting items are required.

Independent verification, where a claim is commercially significant, is available from laboratories such as Intertek, and optical measurement equipment for building a test setup is documented by suppliers such as Keyence.

Turning measurements into a specification

Write the condition, then the value.

A specification item for brightness should read as a value in a condition: luminance at a stated angle, after a stated warm-up, at a stated brightness setting, at defined points, in defined ambient light. The same structure applies to contrast, uniformity and colour. Written that way, the requirement can be verified by measurement rather than by argument, and the supplier’s report can be checked against it line by line.

The pattern is the same one CDTech uses for its own module data, and it is the reason a specification review is part of the quoting process rather than an afterthought. Related reading is collected at the display and touch solutions centre.