22 inch LCD monitors: the 3rd wave! (page 4: Color rendering)

Color rendering

As usual with the help of our probe, we measured the difference between ideal colors and the ones actually displayed on the screen giving us a value called average DeltaE 94. We understand this data from our experience in this domain, however, this isn’t the case for everyone. We therefore are putting into place a new notation system based on our measurements, and which will immediately give you a better notion of quality in this area. Currently, there are no standards for color fidelity and manufacturers can pretty much say and do what they like (see the saga of the Samsung 226BW...). This system will only be for our website but at least it will be a step in the right direction.

By taking the differences between the measured colors and actual gamut found on the screen, a grade from A+ to F- will be given (and less if necessary). Of course, the higher up in the alphabet, the better the color fidelity. An A+ monitor will necessarily be factory pre-calibrated with an average DeltaE of less than 2 and with an extended gamut.

Ouch! Here, we realize that we can’t have all the finest characteristics unified on a single screen.

The Iiyama E2201W is reactive and there is no delay in the display of images, however, pre-set colors are imperfect.

The ViewSonic VX2255wmb offers good color fidelity without adjustment and there is little input lag. However, it is a 5 ms. On the other hand, it compensates for this with high end ergonomics.

Opting for a Samsung 2232BW on the other hand is like playing Russian roulette. Depending on the version, we either have very good color fidelity a B+ for the S model (meaning a Samsung panel) or, and this is rare, an E rating. Besides the outsourced panels that Samsung uses, we really never receive screens with such poor presettings. And however, Samsung assured us that it offers the same quality in both versions. You be the judge.

Fortunately, with the Iiyama by changing brightness to 79, contrast to 30, red to 94, green to 91 and blue to 98 ; the average falls to 3.6 and with a significant improvement to the naked eye. This means the grade goes from a C+ to a B- which is much better.

In the same line of thinking, for those consumers that unfortunately received a C panel in their 2232BW, there is some hope. We now understand why some readers mentioned a blue or red dominance. In fact, black to lighter colors were bluish, while in lighter shades to white they were pinkish. This would be an easy adjustment on a (very) high end model where you can independently adjust dominances; however, on an economical product such as this one, it’s impossible. We ended up playing with auto and manual settings to finally have the best rendering in Personalized mode/Warm Colors. If we take our previous table, with these settings the 2232BW C receives a C+, a substantial improvement.

The table above does not replace the usual detailed explanations. For those who are interested, here is the normal analysis of results.

The color quality is measured with the LaCie BlueEye Pro colorimeter, which in fact, is a Gretag colorimeter coupled to the software suite developed by LaCie.

Just to remind you, we work with a value called DeltaE. It represents a measurement between the color requested and the one really displayed on the monitor. The higher the result obtained, the less true colors are. More precisely, here is how to interpret the graphs:
- Delta E > 3 : the desired color is noticeably different from the one on the screen
- 2 < Delta E < 3 : color quality is satisfactory, but a graphic designer probably wouldn’t be content
- 1 < Delta E < 2 colors are accurate
- Delta E < 1 : the result is perfect

DeltaE by default

You may have noticed that results are consistent with the grades above. There is the ViewSonic in the lead, then the others with an average dE of around 3, followed by those with a value of more than 4. Amongst all of these screens, only the Samsung 226CW is a wide gamut, and this doesn’t really give it any advantage here. Moreover, Samsung doesn’t offer the option of easily switching between wide gamut and sRGB. Colors are partially distorted and, for example, this prevents the proper display of photos. Also, our sensor which is wide gamut compatible failed to calibrate this monitor. Everything was well adjusted except in the extreme reds, which remained too intense.

As for the 2232BW C, with presettings we find its dE in the above table and after manual adjustment this figure decreases to 4.4.

Besides color rendering, there is a point that concerns more and more people, depth of black. We measured a constant brightness of 200 cd/m² in white.

Depth of black (cd/m²) with white at 200 cd/m²

The deeper the black or closer to 0, the better. You will perceive more contrast in images on these monitors. However, color fidelity also has to be good to really take advantage of this. Therefore we had especially good scores for the Belinea and Nec which combine good color rendering and depth of black. The Samsung 2232BW S series does well in this test for the same reasons. Note that the C version also has a good depth in black and besides its poor preset colors, the two versions of this screen are identical as you will see in the following tests. Moreover, some users complained on the forum that the C version had bleeding defects (backlighting which forms a bright halo along the edges). We looked for this on our two 2232BWs one of which was store bought and we didn’t find any problems.

By the way, in response to a few emails on this subject: We systematically measure the brightness homogeneity of screens; however, I’ve taken the liberty to not include this information except in two cases: a particularly good or bad result; and if the screen is destined for professional graphic artists. The reason is that a poor result in this domain can be misleading. For example, you might think that a screen with a 30% difference in brightness would be catastrophic while a 20 % difference would be much better. This only seems logical but is incorrect. Most of the time even expert eyes can’t tell 30% difference is common and unnoticeable. On the other hand, pro screens are often at 15%. There is also another problem with this test given that the result depends on a number of uncontrollable factors. Homogeneity problems can come from too much pressure when closing off the panel, for example (rare); however the most common is from jolts during transport. Panel homogeneity is assured by mirror like reflectors that are in charge of properly dividing up the light from 2 to 16 backlights. Shocks can slightly change their position and therefore their efficiency. This can be in transport from the factory to the ship/truck/airplane, then in the trip to the store, and finally on the way to your desk top, the last two probably being the most critical. So in the end, you can see how our results might not be too representative of the overall product. Anyway, back to the tests!

Most monitors are under the 0.30 cd/m² mark and so are at least satisfactory in this domain. The above measurements result in the following contrast ratios:

Contrast ratio measured at 200 cd/m² in white (xxx:1)

There were no bad results here. All have contrast ratios of more than 600:1 which is very good. This illustrates our opinion on the overall improvements that these monitors represent.

Each time, we analyze 18 standard color patches and 16 of these results are given in the following graphs: