Quote:
Originally Posted by
lgans316 
Purple Tinge is acceptable but not Purple Snakes (thick Purple lines crawling on the fizzing pixels).
I agree! Purple snakes or fizzing pixels along contours is simply a bug in processing. It is not a property of the panel.
To understand how it might happen, you have to realize that the data coming into the panel is YBR data, which must be converted to RGB before driving the panel. The scaling, contrast/brightness, and EE can be done in the YBR, RGB or some proprietary intermediate colorspace.
It is very possible to input a YBR pixel value which is an intense, almost white color, yet mathematically has green hue. In fact, what has happened is that although the hue is green, the intensity brings it close to the white point.
The problem is that if you do the conversion from YBR to RGB, the green value can actually go beyond maximum. On a buggy system, a green 256 on an 8-bit output can roll around to zero, instead of saturating at 255!!
Hence, on a greenish-but-almost-white contour, the color turns purple (actually a magenta).
This is a rookie image-processing mistake. It is a firmware bug -- pure and simple. It should not be accepted by anyone who sees it. Panasonic has acknowledged the problem, and have a fix which involves firmware and a new board (from reading this thread). It has nothing to do with the plasma panel.
Now, purplish tinges in dark areas is a different story, and has more to do with calibration of the various colors at low intensity (perhaps this is difficult to calibrate on plasma?).
Just my $0.02 from an image-processing professional.