The Secret Sauces of THG's LCD Tests

Contrast And Uniformity

To conclude with our set of measurements, we conducted a few interesting static tests from the point of view of an observer. First of all, we measured the stability of the contrast as a function of the brightness. In theory, contrast and brightness are two independent settings and, therefore, when you change the brightness of the screen, the performance in terms of contrast shouldn't alter. Unfortunately, contrast varies between certain devices.

Let's say that you work close to a window, for example. Whenever the sun is shining, you tend to turn up the brightness of the screen to be able to see it better. In contrast (forgive the pun), if you're working late at night or if you're watching a movie in the dark, you reduce the brightness of your screen so as not to tire your eyes. If the contrast remains consistent; no problem. On the other hand, if it changes, then you'll find yourself having to repeatedly adjust your screen, bit by bit, to keep up with the fluctuations in the ambient light. It's not very practical, you have to admit.

So, we measure the contrast for each value of brightness in order to confirm its consistency. This is useful information because it gives an indication of the ease with which you can adjust the device. In addition, if the contrast isn't constant the curve shows us the brightness setting, which gives the best contrast for the user.

The method that we use is still a level more rigorous than that of the manufacturers.

The majority of manufacturers use the Fo-Fo method (Full ON, Full OFF):

• Take a white image, measure the brightness at the center of the screen: ON
• Take a black image, measure the intensity: OFF
• The contrast is then the difference between the two: ON/OFF

This leads to some impressive figures, especially when you take into consideration their unrealism and uselessness. We prefer to use the ANSI method:

• On the same image, we display eight white and eight black rectangles.
• We measure the brightness of each black rectangle, then each white rectangle.
• Then we calculate the average contrast between adjacent rectangles.
• Finally, we compute an average of the contrasts to obtain the ANSI contrast.

This method is much more realistic since, after all, that's what we're interested in: the contrast that is obtainable with the same displayed image. Of course, the values of contrast measured aren't the best obtainable but we've decided on the ANSI method because it corresponds to the real-world situation in which the monitor will be used.