LCD monitor burn-in, revisited (Shallow Thoughts)

LCD monitor burn-in, revisited

The problem I noticed back in March was that my xchat window, always positioned at the same spot on the screen, was getting burned into the monitor. Against a blank white or grey screen, I could clearly see where that window's titlebar normally was, and rows of horizontal lines where the text was.

First, I changed my ways to avoid having windows in the same place all the time. I changed my window manager settings to remove most window placement settings, I removed directives to show any windows on all desktops, and a worked on developing a habit of moving windows around periodically to slightly different locations (I think I'll have my Firefox window on the upper right this afternoon). I don't really like that -- I guess I'm enough stuck in my ways that I like knowing that I can look to the upper left for web pages and the lower right for IRC -- but it's not that bit a deal.

And, happiness, my burned in xchat lines went away. My old bad behavior had not permanently burned in the pixels on my nice monitor.

But that's only part of the story -- because if you look at the photo from March, xchat is not all you see that's burned in. There's also the wavy stuff going across the lower 1/4 of the screen -- and that didn't correspond to any window I'd been running.

I thought maybe it was left over from some Windows wallpaper used by the monitor's previous owner. But none of the standard wallpapers on my Vaio's WinXP partition have this pattern. (What it reminds me most is the data I used to analyze from a cell-sorting machine in my first computer job. Somehow I suspect that's not the culprit.)

These patterns, unfortunately, are not going away. In fact, the ones along the top and bottom left edges are pretty clearly getting worse, and eventually, alas, I'll probably have to replace the monitor. But meanwhile, they vary a lot.

When I first turn on the monitor in the morning, most of the time I can't see the burn-in at all. After a long day of use, it's usually pretty obvious. In between, though, there's huge variation. Sometimes they appear after an hour of use; sometimes I can go most of the day before the burn-in starts becoming visible.

It doesn't seem to be particularly temperature sensitive, and it doesn't seem to vary much with which background image I'm using that day. Sometimes when I'm going to be away for a while, I display an all white screen -- I've read a few reports indicating that can help, and anecdotally I think it does. I should probably keep better statistics on temperature, background color and time to find out what's really affecting this. Maybe I could use it as a homework project in the new Linuxchix R/Stats course!