LCD Panels Need More Class

By Alfred Poor

Information Display, June 2004 - This column originally appeared in the June 2004 issue of Information Display magazine, and is used with permission.

A customer shells out good money for an LCD monitor (or a notebook with an LCD screen), turns it on, and discovers a couple of tiny colored dots that continue to glow even when the screen is supposed to show black. The immediate reaction is "I've got a defective unit," and the user tries to exchange it for one without defects - only to be told by technical support that such defects are normal with LCD panels and the company will not exchange it. The user is unhappily surprised.

The good news is that product manufacturers now are making more of an effort to explain their pixel-defect policies up front to consumers, in the user manuals and on Web sites. The bad news is that it is not always easy for end users to understand. As an example, the following policy from an unnamed company was selected at random.

"The LCD monitor shall not show more than (whichever of the following limits is reached first):

• A total of eight non-performing pixels (of whatever type), or

• Five bright non-performing pixels appearing as a red, green, blue, yellow, cyan, magenta, or white dot on a dark or black background, or

• Five dark non-performing pixels appearing as a black dot on a bright or white background, or

• Two non-performing pixels of any type located less than 10 mm from each other."

If you are not a display-industry insider, I doubt that this will make much sense to you. And how do you compare this to the claims of other LCD manufacturers?

Another manufacturer claims in its end-user manual that the panels used in its 1280 x 1024-pixel monitor are at least 99.999 free of stuck pixels. But given the 3,932,160 subpixels on the panel (a number that they also cite in the manual), this still means that there could be as many as 39 stuck subpixels and the monitor would still comply with this quality statement.  Would an end user understand this enough to be able to compare the quality policies of these two companies?

When we buy beef or butter at the grocery store, we can see right on the outside of the package whether the product is Grade A or Choice or something else.  Why can't LCD-panel products have similar consumer labeling so buyers can compare quality standards more easily? What we need is a consumer-oriented classifying system that product makers can use to advertise the quality of the panel used in their display products.

But wait! Such a system already exists. The ISO 13406-2 standard specifies a variety of attributes for LCD panels, including a classification system for pixel defects. Class I panels have zero defects, period. Class II panels quantify the number of permissible defects per million pixels. Qualifying panels can have up to two bright pixel defects, two dark pixels, and up to five subpixel defects, either bright or dark. Two or more defects within a 5 x 5-pixel matrix are not permitted.

The standard also sets specifications for Class III and IV panels. Class III units can have up to five bright pixels, 15 dark, and 50 subpixel defects per million pixels, and up to five subpixel clusters in the 5 x 5-pixel-matrix range. Class IV permits hundreds of defects per million pixels.

The ISO 13406-2 standard is better known in Europe than in the United States, but many LCD-monitor vendors acknowledge it. For example, NEC-Mitsubishi asserts that all their LCD-monitor products meet or exceed the Class II standard for pixel defects. Other manufacturers are less specific, simply claiming that their products are "ISO 13406-2-compliant," which does not tell the buyer much about their pixel-defect policy.

The ISO standard has a number of limitations. It does not explain whether the permissible detects are prorated: Does a panel with 1.5 million pixels get one-and-a-half times the defects allowed for Class II qualification? Also missing are details on precisely how the defects are to be defined and measured. The biggest problem of all, however, is that LCD-panel production has outpaced the standard, and now the granularity of the classes is too coarse to be of any use. Few buyers can afford the cost of a guaranteed defect-free Class I panel, yet few would be satisfied with anything less than what Class II permits. 

Another factor is that unlike a rose - which is a rose is a rose - a defect is not always a defect. As pixel counts increase and pixel pitches decrease, a single pixel defect can be more and more difficult for the end user to perceive. A single stuck pixel on a small high-resolution panel might be more difficult to spot than would a stuck subpixel on a larger lower-resolution panel. So, perhaps a classification system should take this into account.

But who can come up with such a system?  Fortunately, there are now two groups actively grappling with related issues. One is the Standard Panels Working Group (SPWG) (www.spwg.org), which published a draft of their 3.0 specification in December 2003.  This specification does address a wide variety of pixel defects, but fails to quantify any of them. The other group is the new VESA Panel Committee (www.vesa.org/membership_committees_Panel.html), which grew out of a Special Interest Group (SIG) that was started in 2003. This group is focusing on mechanical issues such as standardized mounting holes and other dimensions, and does not address defects at all.

Both of these groups are in a position to tackle the knotty problem of making pixel-defect policies more understandable - and palatable - to the end-user buying LCD-panel products. The SPWG has a leg up on the process by defining different types of visible defects in their existing draft standard. The VESA committee has a longer way to go, but the organization has a proven and trusted track record in sorting out nasty technological problems for the direct benefit of buyers as well as the benefit of the display industry as a whole. Remember how they solved the SVGA timing issue, or the problem of mounting LCD monitors on aftermarket stands, or the potential Balkanization of the digital display interface. VESA has the clout and capability to solve this problem, if it has the conviction.

I recognize that this is not a simple matter.  The quality standards specified in contracts between display manufacturers and their suppliers are complex and sometimes a matter of proprietary interest. These manufacturers may resist being more forthcoming to consumers about just how many bad pixels it takes before they will deem the whole product defective.

But this is an important issue for consumers, and it's only about to get worse. So far, most of the panels have been sold in notebooks and computer monitors, in markets where the buyers are fairly savvy about technological issues. Yet even these people are confused and frustrated by the arcane pixel-defect policies that they encounter. What's going to happen when the buyers are less knowledgeable, but investing larger sums of money in each purchase and viewing the images for more than just spreadsheets and Web browsing? The LCD industry is banking on growth from television sales, and I guarantee that these buyers will be less tolerant of and understanding about defects than your average computer user.

The time to provide consumers with a useful classification system of LCD products based on pixel-defect policies is now, before the flow of complaints becomes a flood that damages the reputation of the entire industry.