Display Testing Explained: How We Test Monitors and TVs

Our display benchmarks can help you decide what screen to put on your desktop and in your living room. So, in order to provide a little greater insight into our monitor and HDTV testing methods, we’re running down our procedures step-by-step right here.

In every Tom’s Hardware monitor or HDTV review, we briefly describe our testing methods. Our goal is to perform a series of benchmark tests that look at each aspect of video performance in order to help you decide which display is best for your particular application.

We separate the tests into six major categories: contrast, grayscale, gamma, color, screen uniformity, and panel response. In doing this, you can prioritize image parameters and decide which is most important before you make your purchase.

In this detailed rundown, we’ll describe our testing methods, what equipment we use, and what the data means in terms of image quality and display usability.


To measure and calibrate monitors, we use an X-Rite i1Pro spectrophotometer, Spectracal C6 colorimeter and version of SpectraCal’s CalMAN software. Test patterns are provided by AccuPel DVG-5000 and DVDO AVLab TPG video signal generators via HDMI. This approach removes video cards and drivers from the signal chain, allowing the display to receive true reference patterns.

The i1Pro is very accurate and best-suited for measuring color on all types of displays, regardless of the backlight technology used. Since the C6 is more consistent when measuring luminance, we use it for our contrast and gamma tests.

The AccuPel DVG-5000 is capable of generating all types of video signals at any resolution and refresh rate up to 1920x1080 at 60 Hz. It can also display motion patterns to evaluate a monitor's video processing capabilities, with 3D patterns available in every format. This allows us to measure color and grayscale performance, crosstalk, and ghosting in 3D content via the 3D glasses.

A DVDO generator is the latest addition to our lab. It supports resolutions up to 4096x2160. We’re using it to verify the proper signal handling of QHD and UHD displays.

On rare occasions, a monitor isn’t compatible with either the AccuPel or the DVDO generators. Then we connect it directly to a PC and use Spectracal’s CalPC Client to generate patterns. All look-up tables are disabled so we can evaluate the product’s raw performance. Calibration is still performed with OSD controls-only. Direct Display Control is not used unless there is no other way to correct errors.


The i1Pro or C6 is placed at the center of the screen (unless we’re measuring uniformity) and sealed against it to block out any ambient light. The AccuPel pattern generator (bottom-left) is controlled by CalMAN through USB, which is running on the Dell XPS laptop on the right.

Our version of CalMAN Ultimate allows us to design all of the screens and workflows to best suit the purpose at hand. To that end, we’ve created a display review workflow from scratch. This way, we can collect all of the data we need with a concise and efficient set of measurements.

The charts show us the RGB levels, gamma response, and Delta E error for every brightness point from zero to 100 percent. The table conveys the raw data for each measurement. In the upper-left are luminance, average gamma, Delta E, and contrast ratio values. This screen can also be used for individual luminance measurements. We simply select a signal level at the bottom (0 to 100 percent) and take a reading. CalMAN calculates things like contrast ratio and gamma for us.

Every primary and secondary color is measured at 20-, 40-, 60-, 80-, and 100-percent saturation. The color saturation level is simply the distance from the white point on the CIE chart. You can see the targets moving out from white in a straight line. The farther a point is from center, the greater the saturation until you hit 100 percent at the edge of the gamut triangle. This shows us the display’s response at a cross-section of color points. Many monitors score well when only the 100-percent saturations are measured. Hitting the targets at lower saturations is more difficult, factoring into our average Delta E value (which explains why our Delta E values are sometimes higher than those reported by other publications).

In the following pages, we’ll explain each group of tests in greater detail. Let’s take a look.

Create a new thread in the UK Article comments forum about this subject
This thread is closed for comments
No comments yet
Comment from the forums
    Your comment