Wednesday, December 13, 2006

A New Bedroom HDTV for Me, Part III

In A New Bedroom HDTV for Me, Part II I complained about a minor defect in the Sony KDL-40XBR2 40" LCD panel I bought for my bedroom: "an irregular swath, predominantly vertical, sitting slightly to the left of the center of the screen in roughly the top two-thirds of the image." A portion of a solid gray test pattern showed up as slightly too dark in the swath in question.

Further experimentation now reveals the swath is actually more regular than I originally thought. It runs from the top to the bottom of the screen and seems to be part of a series of other vertical brightness "troughs" and "crests" that flank it to either side. It looks like the supposedly perfectly flat LCD panel is faintly "crumpled," in the way a rectangular rug that has been bunched up in one direction can look crumpled .

Another way to describe it is that there is a subtle "accordion" effect in the LCD panel. The "pleats" are, however, not sharp-edged but gradual.

The main "trough," or "accordion pleat," the one almost at screen center, even has a very slight pinkish tinge, where the test pattern is supposedly a uniform light gray.

I've managed to discover that the type of defect I have in my Sony is one that is well-known to afflict LCDs. It is called mura, the Japanese word for blemish, unevenness, or inconsistency. This PDF gives some technical insight into the phenomenon, as do this one and this one. The image at right was taken from the first of these three. It shows one of the several kinds of mura defects in LCDs, the so-called vertical-band or v-band mura. Notice the darker vertical stripe a bit to the right of center.(Other mura types include cluster muras, rubbing muras, and light-leak muras.)

The v-band mura shown here resembles the main brightness "trough" on my Sony screen, except that mine is slightly to the left of center, while this one is slightly to the right. (This example also manifests a central, over-bright "hot spot." I assume that this is an artifact in the example graphic and is not relevant to the overall mura defect issue. My Sony certainly displays no such hot spot.)

This example doesn't show the same flanking "troughs" and "crests" to either side of the main "trough" as my Sony exhibits. Nor does the example demonstrate the ultra-slight pinkish tinge near my main "trough." Yet the example above gives you a pretty good idea of what you would see if you looked at my TV while it was rendering a light gray test image.

Breaking out my digital camera, I took this picture of my Sony displaying a solid gray 30 IRE test screen. (The IRE scale from 0 IRE to 100 IRE measures shades of gray from pure black to full white. 30 IRE is a dark medium gray.) There is a vertical stripe just a tad to the left of the center mark as indicated by the Sony logo. There is lesser non-uniformity of illumination (fainter stripes, "crests," "troughs") surrounding it, all vertically oriented. In fact, you might consider the relatively bright "wings" of the image to be overly illuminated, as is a "crest" to the right of the center line.

In this picture you can't see the pinkish color tinge that I see with my eyes, the one associated with the main vertical band or "trough."

While I was experimenting with the above, I chanced upon the gamma test pattern of the Avia DVD and decided to play around with the Sony's gamma setting. I don't have a lot of confidence in this test pattern giving accurate numbers with non-CRT displays ... but for what it's worth, with the TV's gamma setting Off, the Avia gamma figure read 2.2. Each "higher" Sony gamma setting — there are four choices — decreased Avia gamma by 0.1. So at the maximum setting, Avia reported a gamma figure of 1.8.

Gamma affects the apparent contrast in the image. The higher the gamma figure — the "lower" the Sony setting for gamma — the more contrast there appears to be. When gamma is 2.2, there appears to be more contrast than when gamma is 1.8. 2.2 is considered fairly standard ... though, again, I won't vouch for the precision of the Avia numbers on an LCD TV, or on any TV where "gamma" is generated digitally and not an intrinsic function of a cathode ray tube that displays a picture using interlaced scanning.


Anonymous said...

I have similar problem with my Sony SDM-E96D computer LCD display. The difference is that I have both vertical and horizontal bands, crossing at the center of the screen, forming a cross. Here is a image:

eric said...


Thanks for sharing.