Originally Posted by Elemental1
Those screenshots don't look subtle to me.
They aren't meant to be subtle. They are closeups taken so that the problem becomes obvious. Unfortunately taking a digital photo of a plasma screen from that close up also introduces several other artifacts which make the photos look very ugly and also look like the screen is exhibiting problems that aren't really there in real life.
If you get your eyes within a few inches of a plasma it's not going to look pretty. You'll see the pixels, you'll notice that the plasma tubes flicker a bit because they are being driven by AC signals and by and large the colors will wash out as your eye loses a reference to compare them against. It won't be the pretty image that you are used to viewing from a few feet away.
If you put a digital camera within a few inches of a plasma it will look even uglier. A CCD sensor is not designed for the task at hand and will see not only the problems that your eyeball will notice that close, but others as well due to the nature of the sensor vs. what it is being fed. It will have trouble focusing, the CCD will be experiencing contrasts that are way outside of its intended usage, it won't be able to determine a good white balance, and it will be trying to capture it in a color space that doesn't match the light that is fed to it.
But, while all of those problems make the images look flat, blurry and washed out and make it look like the set has poor PQ, they are anomalies of the way that the images are captured and they are not the problem we are seeing and discussing here. We are talking about the bands of magenta that appear where they shouldn't in those pics.
If you look at the closeup on the forest, the one that looks like a lava lamp of green and yellow - it's ugly and that ugliness is an artifact of the image capture - but even through the ugliness you should only see shades of green and yellow in the ugliness. But instead there are patches of purple mixed in. Where did the purple come from? It came from a mismatch in the response curves of the primary colors.
If you look at the picture of the man standing against the forest background - sure he is out of focus and the colors aren't that great. But if you look off in the forest part of the picture you again see the purple. But at this distance it is harder to make out. Once you see it in the "green and yellow lava lamp" closeup of the forest, it will be easier to spot in that photo too.
Again, behind Kong there is a green patch on the background that should be all green, but has a strange band of purple in it. Sure, the shot on the whole doesn't have very good color, but that is because the lighting conditions weren't conducive to a good digicam shot. Either way, though, the purple hue is just plain out of place no matter what the exposure problems may have been.
Edit: Also, has anybody bothered to not test the cablebox and plug the RF in directly to the set?
The magenta/purple patches of color appear regardless of source. It's related to what is in the image, not the quality of the signal. If the signal were improved or degraded then you'd just get a prettier or an uglier picture of the purple bands.
This isn't a "bad signal" problem and it isn't related to the fact that those screenshots, on a whole, look crappy. They are crappy pictures of a very real problem that becomes more subtle as you move away from the set. (Edit: I wanted to reiterate that the "crappiness" comes from the near impossible conditions of the shot, not from a failing of the photographer...)
Look, I work in digital imaging for a living. I've seen these color bands before. They happen when the display device you are using has odd response curves for the primary colors that you didn't notice or that you forgot to take into account when processing the image for display or when you tried to take them into account but got the parameters of the adjustments wrong.