Two simple ideas
Brightness is how much light a screen gives off. It is usually measured as luminance, expressed in candelas per square meter, often called nits. A higher number means a brighter image. Contrast describes the gap between the brightest white and the darkest black the screen can show at the same time or under the same conditions.
These two qualities are linked but not the same. A screen can be very bright and have weak blacks, or be fairly dim and still look rich in a dark room. What matters is the combination and the surroundings.
What a contrast ratio is
A contrast ratio divides the luminance of white by the luminance of black. Here is a hypothetical example, with made-up numbers chosen for easy arithmetic. If a screen’s white measures 200 nits and its black measures 0.2 nits, the ratio is 200 divided by 0.2, which is 1000, written as 1000:1. If black were 0.1 nits with the same white, the ratio would be 2000:1.
Two things follow. First, the ratio is very sensitive to black level, because the black number is small and is the divisor. A tiny change in black makes a big change in the ratio. Second, a quoted ratio says nothing about how bright the screen is. Two screens can share the same ratio while one is much brighter.
Quoted figures can also differ in how they are measured. Some are static, meaning white and black are measured on the screen under fixed settings. Others are dynamic, where the screen changes its backlight between scenes, which can produce larger numbers that do not describe a single image. When comparing, it helps to know which kind is stated, a point covered in the article on reading display specifications.
How room light changes the result
A screen does not show black by emitting less than nothing. It shows black by emitting little light, and the room adds light that reflects off the surface. In a bright room the reflection adds to both white and black, but the effect on black is proportionally much larger. The perceived contrast drops.
A hypothetical example: say a screen’s black is 0.2 nits and a bright room adds 2 nits of reflected light to its surface. The effective black becomes 2.2 nits. If white is 200 nits, adding the same reflection makes it 202. The ratio becomes 202 divided by 2.2, about 92:1, far lower than the 1000:1 in a dark room. The numbers are invented, but the pattern is real: ambient light can reduce contrast a lot.
This is why the type of surface matters. Glossy and matte finishes treat reflections differently, which is explored in the article on screen coatings. It is also why screens that adjust to ambient light, discussed in the article on ambient light sensors, can feel more comfortable and more legible.
Checking brightness and contrast by eye
Simple visual checks can reveal a lot, as long as the room is controlled.
- Dim the room and let the screen warm up for a while. Many displays shift slightly during the first part of an hour.
- Turn off automatic features such as dynamic contrast and automatic brightness while testing, because they change the image as you watch.
- Look at a black image in a dim room. A very dark gray glow means the black level is weak. On an LCD, uneven glow at the edges may be backlight bleed.
- Look at a white image. It should be clean and even, without a strong tint.
- Look at a pattern of near-black and near-white steps. If the darkest bars merge into black, shadow detail is being lost. If the brightest bars merge into white, highlights are being clipped. Test patterns are explained in their own article.
- Repeat in the room where the screen is actually used, since that is what matters in practice.
Your eye adapts, so judging absolute brightness is unreliable. Comparing two screens side by side, or the same screen at two settings, is much more dependable than judging one alone.
General measuring approaches
When numbers matter, people use a light-measuring instrument placed against or pointed at the screen. The general method is:
- display a full-screen white image and measure its luminance
- display a full-screen black image and measure its luminance
- divide white by black to get a ratio
- repeat at several points on the screen to see how even it is
A more demanding approach uses a checkerboard of black and white squares, which tests contrast while both extremes are present at once. Results can differ from the full-screen method, especially on screens whose backlight or pixel behavior changes with image content. A screen that lights pixels individually and can turn them fully off will often behave differently from one with an LCD backlight.
The instrument needs to suit the screen type, and measurements are repeatable only when the room, warm-up time and settings are kept the same. This article does not recommend particular instruments, and any figure from a quick home measurement should be treated as approximate.
Brightness and settings
Brightness setting is not a test result. A brightness slider usually changes backlight output or pixel drive, and the setting numbers do not correspond to a fixed luminance. For a rough rule, many people use lower brightness in a dim room and higher brightness in a bright one, which also relates to eye comfort. A setting that makes white painfully bright in the evening can be too dim in daylight.
Peak and sustained values also differ. Some screens can reach a high level on a small bright area but cannot hold it across the whole screen, which is the subject of the article on peak brightness versus sustained brightness. Related ideas appear in high dynamic range, which asks screens to show a wider span between dark and bright.
Things that skew a test
- Reflections from windows or lamps in the screen
- Viewing angle: some panel types lose contrast when viewed off-center, covered under viewing angles
- Picture modes that boost or limit brightness
- Aging: over time, screens can dim and shift, as the article on panel aging describes
- Content: automatic dimming can change the readings
Keeping a note of conditions makes comparisons more meaningful. A good result for one purpose, such as reading in a bright office, is not necessarily a good result for another, such as viewing a film in a darkened room, and no single number answers both.