12. Image Quality, Screen Brightness And Contrast

We assess laptop and notebook screen brightness using a luminance meter, which measures the intensity of light emitted by an object. It uses a black tube pressed against or held very near to the object being measured to block out all or almost all ambient (room) light. We let test computers run for 30 minutes before beginning luminance tests; this gives the screen backlight enough time to warm up and rise to the brightness level we have set for the computer.

Our Gossen luminance meter has its own memory and connects to a computer so data can be downloaded easily rather than manually copied.

Luminance meters measure brightness in candelas, more specifically candelas per square meter, a unit also called a nit. Our battery life tests require that each mobile computer's brightness be set to between 60 and 70 nits using the luminance meter, a setting considered adequate for viewing an LCD screen in most moderately lit indoor environments.

An ideal laptop or notebook would have the same brightness across its screen, which in technical terms means the screen's spatial uniformity of brightness would be perfect. However, both the screen itself and the backlight that illuminates the screen are not perfect, which leads to measurable imperfections in the spatial uniformity of screen brightness across the panel.

To test the brightness and spatial uniformity of brightness of a laptop's or notebook's screen we use an image that divides a white (RGB 255, 255, 255) screen into 64 rectangles. We resize the image so that it fits as closely as possible on the test computer's screen. Next, we set brightness to full and measure and record the brightness of each of the rectangles with our luminance meter. We always measure screen brightness with the test computer running on AC power.

We calculate screen brightness as the average of the 64 measures, and compute spatial uniformity of brightness as the standard deviation of the 64 sample measures. A standard deviation of 0 would mean that spatial uniformity of brightness across the screen was perfect. The higher the standard deviation of the sample measures, the lower the spatial uniformity of brightness. However, except with very high quality screens, the brighter the screen, the higher the standard deviation will tend to be. So, you should take brightness into account when interpreting standard deviation.