Color gradients are reproduced well. There appears barely visible banding at lower values of contrast, but it is really hard to discern even in the special test image.

The monitor’s color gamut is typical for a model with old backlight lamps: it coincides with the sRGB triangle in blues, smaller in reds and larger in greens. There are monitors with new backlight lamps that use improved phosphors and ensure a larger color gamut in greens. Red and blue remain unchanged, though.

I want to remind you that the monitor’s color gamut is not a measure of the precision of its color reproduction. It only indicates how pure its colors are. On a majority of modern monitors (excepting those with LED-based backlight), red has a slightly yellowish tincture. They just cannot display a truly deep red.

As for color reproduction in general, the monitor’s ability to combine the three basic colors into halftones is more important than the basic colors proper. Particularly, this ability is described by the so-called gamma curves: the graphs that show the dependence between the real screen brightness and the numeric value at the graphics card’s output. For a correctly set-up monitor this graph is a power function with an exponent of 2.2. A distortion in the shape of the gamma curve means that the monitor displays some halftones brighter or darker than they should be. A difference between the curves of the different basic colors means that some halftones have an undesired tonal coloring.

The gamma curves of the 2263DX are only good. The blue curve is lower than necessary, and blue halftones are displayed darker and with more contrast than they should be. The red and green curves are closer to the ideal one, but have a small bend in the top right part of the diagram. It indicates that the lightest halftones of these colors won’t be distinguishable from each other. These drawbacks aren’t conspicuous, though.

The curves improve at the reduced brightness and contrast and now go close to the ideal curve. It’s rather typical of many monitors: they have a slightly higher value of contrast than necessary at the default settings, which leads to color distortions. Well, modern monitors are usually so bright that no one works with them at the default settings anyway.

Our calibrator records the curves of different colors and normalizes them (so that they go from point {0, 0} to point {255, 255} of the diagram) individually and the resulting diagram does not accurately represent the deviation of the curves from each other, which results in the above-mentioned toning of certain halftones. The diagram gives you a qualitative but not quantitative representation. To get the latter, we measure the color temperature of gray in four points, from dark gray to pure white, and compare the results. Ideally, if there is no parasitic toning of halftones, gray looks gray (but not reddish or bluish) irrespective of its brightness. That is, all the four measurements must yield the same result for every setup variant.

The SyncMaster 2263DX is somewhat below average level in this respect. In every mode available in its menu there is a 1000K and higher difference between the temperatures of different grays. For TN-based monitors, I consider a temperature dispersion of 1000K and lower as average and within 500K as good.

This ends our tests of color reproduction. As I have found out, the SyncMaster 2263DX is average in its class: a standard color gamut, good but not ideal gamma curves, and an acceptable quality of color temperature setup.

Our next test measures the most popular parameter of an LCD monitor, its response time. We use the GtG method (see the beginning of this section), calculating the average of all possible transitions between different levels of gray from 0 (black) to 255 (white) stepping 32. The switching of the pixel’s state is recorded by an oscilloscope connected to a photo-sensor attached to the monitor’s screen. Below is the diagram that shows the duration of each transition.

The response time average (for all the columns of the diagram) is 15.2 milliseconds (GtG). It is three times as high as the specified value of 5 milliseconds. Why? Because the measurement methods differ. The official 5 milliseconds is the sum of two transitions, from black to white and back into black. As you can see in the diagram above, the columns corresponding to the transitions into white (the farthest row, painted yellow) and into black (the nearest, red, row) are the lowest, i.e. the fastest.

Of course, counting in all the possible transitions in the GtG method is closer to reality than the black-white-black measurement in the ISO method. Do you often play games whose picture consists of black and white pixels only? Thus, the 2263DX is not actually a fast monitor, just like the other models with a specified response time of 5 milliseconds we have tested so far.

There is an opinion that any response time below 16.7 milliseconds is enough for a refresh rate of 60Hz. This is not true. Response time and refresh rate are completely different parameters. It doesn’t matter if the monitor has changed the picture before the arrival of the next frame – this has no effect on the fact that the pictures from two adjacent frames will be displayed both together during the time determined by the matrix speed. If the response time is higher than 16.7 milliseconds, there will be three, not two, pictures displayed at the same time with the arrival of the next frame, and this won’t make the ghosting effect look any better.

Another important parameter of each LCD monitor is the uniformity of its brightness. It may be different for white and for black, so we measure it two times using a photo-sensor whose position relative to the screen is changed with a step of 3 centimeters. The sensor’s reading in each point is entered into a table. The table is used to build a diagram that shows how the monitor’s screen looks in reality. Take note that it is a picture, not a photograph! Its colors are not the real colors of the monitor.

The 2263DX doesn’t have obvious bright spots on black: there are darker bands along the sides of the screen and a rather uniform middle. To be specific, the average difference between the levels of black is 5.6%. The maximum difference is 16.8%, which is quite a good result.

On white, there are darker areas along the edges of the screen, especially on the right, but the rest of the screen is uniform. Thus, the average uniformity is rather low at 5.1% while the maximum deflection is high at 22.8% (it is obviously due to the top right corner of the screen).