Quote:
Originally Posted by
WilliamG 
Quick question: What do you guys do about the color uniformity on your PT-AE8000U? I'm being driven a bit mad by the greenish tone to the middle of image, and I can't see to get rid of it. I notice it on sky-scenes, on anything "grey" on screen etc, and it bugs me to no end. I'm getting kind of sick of it, but have no idea how to remove the green from the image... *sigh*
Any help appreciated....
I think that the degree of color uniformity is determined by A) convergence and B) grayscale calibration. On this projector, I suspect that it's the optics (prisms, lenses, filters, etc.) that are the limiting factor in perfecting convergence. I've taken some convergence photos of my PT-AE8000U to illustrate why I think this.
The first picture shows nearly perfect vertical convergence. The entire screen looks like this top-to-bottom and left-to-right:
Vertical Convergence
Horizontal convergence is a different story. The following three images show the horizontal convergence starting on the right side, moving to the middle-left side, and then to the left side.
In the first picture of the right side, notice how the blue pixels start off with 1-pixel misconvergence toward the left of red and green:
Horizontal Convergence, Right Side
In the second picture of the middle-left side, the blue becomes perfectly converged with red and green.
Horizontal Convergence, Middle-Left Side
In the third picture, the blue shifts to a 1-pixel misconvergence toward the
right of red and green.
Horizontal Convergence, Left Side

So moving across the screen, blue tends to stray inside of red and green. It's a smooth and gradual 2-pixel shift in blue from right to left.
In my thinking, this is caused by the optics imparting differing degrees of light bending depending on the color passing through them. I think this primarily affects horizontal convergence because I believe the three colors of light are split horizontally by prisms in the projector, sent through the three LCD panels, and then re-combined through prisms on their way out the main lens. The light is not subjected to different paths in the vertical direction however.
To improve the color uniformity without doing anything to the convergence, I did a DIY grayscale calibration using an X-Rite Display color sensor and the free ColorHCFR hobbyist calibration software. Once you have tailored a proper individual gamma correction curve on all three colors, red-green-blue will come up out of black heading toward white with almost perfectly equal luminosity as the picture brightens. Without proper gamma correction, at various picture brightness levels there will be incorrect mixtures of the three colors, which will tend to worsen the color uniformity issues that result from the previously mentioned color-dependent differences in how light passes through the optics.
Calibration went a long way toward improving the color uniformity on my projector. It will never be perfect but it is at the point now where it is not really noticeable on actual program material like clouds or white buildings.