Panels that are produced too economically?

Panels that are produced too economically? - BeHardware
>> Monitors

Written by Vincent Alzieu

Published on September 18, 2007

URL: http://www.behardware.com/art/lire/681/

Page 1

Panels that are produced too economically?
Panels that are produced too economically?
We are still reaping the results from the tests and unveiling of the various Samsung SyncMaster 226BW versions. At first, it enabled us to see the number of different panels used for the same reference number as well as the variation in settings. This especially resulted in random color fidelity despite rather similar components.
Another point that has now come up is that for same family of panels from the same manufacturer (Samsung, in this case), colors may appear homogenous to the eye but there are darker zones as soon as we move our eyes from the center of the panel. From what we’ve gathered, this has nothing to do with a potential Mura effect and is more a problem related to the filter applied to panels and viewing angles that are much too small.

Once again the alarm was sounded on the Forum and this brought our attention to the problem. Readers started complaining of darker zones on their screens and imperfect color homogeneity. More specifically, unified images were darker in the upper area of the screen than on the lower area. We didn’t really understand (or notice) this before because none of the monitors that passed through our lab displayed this characteristic.

Then we received the Iiyama ProLite E2201W, a 2 ms TN equipped with a Samsung panel. Reactivity, input lag, and homogeneity measured by our sensor – all matched up perfectly to characteristics of the Samsung SyncMaster 226BW “S” series that we obtained. However it even if you sit directly in front of the screen, it is possible to see tint variations in the corners.

Here is a graph displaying homogeneity measured with our sensor:

More specifically, the maximum brightness difference between two points is 25 % with an average of 8.6%. This is completely within the norm and even better than average. A classic screen can easily go as high as 30 % between two points while the more homogenous ones are at 15 %.

And yet when we look at the center of the panel and simply move our eyes to a corner, the image immediately turns darker:

This is a photo of the panel superimposed on a shot taken directly in front of the screen.
We could start to suspect a defect related to the Mura effect, in other words, poor brightness homogeneity due to a problem with the light reflectors. However, the difference measured between the center point below and the one above was only 4%. Another troubling fact, and which definitively ruled out the Mura effect, was that if we then looked up at the upper right hand corner of the screen, here is what we obtained (the two images are the closest reconstructions we could manage with Photoshop) :

Another possible variation was that if we barely change our gaze the dark zone moved:

We noticed this defect on the first Iiyama ProLite E2201W and then in the company of their representative with a second one that they brought us for verification. We expressed our surprise to them and then to our even greater surprise (and to their credit, we suppose) they admitted to be aware this problem. They didn’t see that the defect wasn’t homogenous and thought it to be associated to a Mura effect. For them, this was already noticed on a more basic product also equipped with a TN panel, that time an AU Optronics.

Page 2

Technological level: the origin of the problem
On a technological level, where does this problem come from?
We have to go back to LCD basics and determine from what layer this defect can originate. By searching the internet, here is the most complete diagram we could find:

However, by the process of elimination, we could quickly rule out the layers shown on this diagram. By directly facing a point on a screen, this assures that polarizing filters are doing their job, liquid crystals are properly turning in all zones, and that backlighting is emitting light in a homogenous manner...

We therefore had to look elsewhere.

Actually, all the diagrams found on line corresponded to TN screens. However, for several years now this technology has evolved to become TN + Film, the Film in question serving to artificially increase viewing angles from 90° to more than 150°. But it is difficult to find illustrations that show their presence! In trying to learn more about them, we actually found the specific name of this innovation to be Twisted Nematic + retardation film, and that one of the most recent additions to TN panels consists of covering it with a "Beam Steering optical Film" (BSF), which fractions light beams in order to open up the viewing angles.

When we know where to look, everything falls into place, however, this took more some searching...

During the 21st International Liquid Crystal Conference which took place in July 2006 in Keystone, Colorado, one of the presentations was in fact devoted to the film in question. This was based on the work of LG-Philips labs with the help of two Korean universities. The principle, however, seems to apply to all modern TN panels and the diagram is meant for the more technologically proficient (please excuse the lack of simple explanations here!). The diagram doesn’t have as many colors and it has to be interpreted, however, in the end we can pick out the “BSF” layer and then determine its repercussions.

The addition of the BSF layer over the final polarizing filter, first of all, results in significantly reducing the amount of light captured compared that in the cells. In this precise example, on the panel surface we go from 264 candelas / m² to 140 cd/m² ! In compensation, the zone on which the technical teams deem the image "good" increases more than 25 %, with the measured contrast resulting in lateral angles going from 50° to 80° and vertically from 20° to 60°. If we add the upper/lower and two lateral angles this gives us a final figure of 160° / 120°, which indeed corresponds to what we measure on current TN panels. (Their measurement is more stringent and, in our opinions, more reliable than the ISO standard, which recently comes to 170°/160°).

On the other hand, while we can imagine that the presentation tried to show the filter in the best possible light, we can see on the diagram that the viewing angles aren’t entirely homogenous. As you can notice, the curve isn’t perfectly symmetrical. On the panel, this will be barely noticeable, but evidently color rendering will slightly vary when looking at different areas on the screen.

In our opinion, the problem found on the Iiyama screen indeed seems to be related to this filter. Moreover, we have come to conclusion that due to economical reasons in the noble effort of reducing production costs (and the final price to consumers), panel manufacturers have put poor quality films on certain more economical panels. This can result in serious problems in light diffraction.

We have started to mention this problem to manufacturer engineering teams. This seems like a logical and coherent course, and we are waiting for responses from teams in Korea, China, and Japan to address this situation and give us their input.
Our work cut out for us
Either way, this is something we will give particular attention to in our upcoming tests. While it is rather difficult to detect with simple measurements (a sensor is of little use here), we will have to rely on a simple visual evaluation.

According to mail in forums in several languages, it appears only some TN equipped monitors are affected by this problem of variable brightness homogeneity. This is quite logical in the way that the suspected film is only applied to TNs. Rival technologies theoretically have no need for it because the viewing angles of MVA, PVA, ASV and even more so IPS panels are larger to begin with.