Following are a new set of grids for making geometry adjustments.
Instead of a black background, these new patterns use a 15% gray background to better approximate the average picture level (APL) of video over time. This should slightly improve the accuracy of pincushion, and also horizontal and vertical sizing adjustments for typical video content.
The patterns can also be used for overscan adjustment, if you want a 2.5% or 5% overscan.
(Minor Note: The 2.5% overscan pattern that's divided into 8ths is not quite as accurate for overscan adjustment as the other patterns because there was no way to perfectly divide the width of the image by 8, and leave an exact 2.5% on the sides. Consequently, it's probably closer to 2.75% or 3% on the sides.)
Used properly, the patterns can also function as a crude ambient/surround reference for high-quality home video content (Blu-ray, DVD, OTA broadcasts, etc.). For direct-view displays in the Region 1 US market*, I generally recommend surround levels in the 13% to 20% gray range for best viewing, with 15% being about the average for well-mastered DVDs & Blu-rays. (*Video content in other parts of the world may be brighter.) An easy way to accommplish that is to adjust the Picture/Contrast control on your display so that a full field gray pattern in that range approximates the average surround levels in the room around and behind the TV. Calibration discs can be very handy for this kind of thing. The better ones should include full field gray patterns which increase in steps of 5% or 10%.
There's alot of variation in video content these days though, which is why I try to accomodate a range of adjustment on my TV between about 13% to 20%. Most professionally mastered/color-corrected video content in the US is on the darker side. IMO, a 15% gray is a pretty good target reference for that.
However, games, commercials, music videos, anime (Japanese animation), and older NTSC material can be considerably brighter, and quite a strain on the eyes to watch with such dark surround levels. So something closer to a 20% (or even brighter) gray surround reference would not be unreasonable for such material.
If you also use your display as a computer or web interface, then you may need to accomodate an even wider range of contrast adjustment. The APL of computer content is about 35%, and the sRGB standard (which is basically now the "official" color space of the web) uses a "gray world" assumption for it's viewing environment, which is essentially the brightness of a 50% gray on your screen. For optimum viewing of computer and web content, the Picture/Contrast setting on your display (or the room lighting) should be adjusted so that a 35% to 50% full field gray on the screen is a close match to your room lighting levels.
Again, you should be able to find full field gray patterns like these on many of the popular calibration discs (or you could just make your own with Photoshop or some other paint program). In the final analysis though, your eyes are really the best judge. So if the picture seems too bright (or dark) with the above rules of thumb, then you should turn the contrast down.
"Full-Swing" Vs. "Studio-Swing" Levels
There are two versions of the grid patterns posted below. The first series uses 0-255 "full-swing" RGB levels. And the second series is compressed to 16-235 "studio-swing" RGB levels. For geometry adjustments, they should both work about the same.
If you want to get picky though... then the correct set to use depends on how you display the images on your TV. I usually convert the patterns to a standard MPEG-2 DVD-Video format (which is why they're all in 720x480 NTSC DVD-Video resolution). And the software that I use to encode the files into DVD-Video compliant MPEG-2 automatically compresses the palette in the images to 16-235 video levels (which is what commercial video DVDs use). So I feed the MPEG encoding software patterns and images authored in 0-255 full-swing levels to ensure proper output levels on my TV. Other MPEG/DVD authoring programs may work differently though, and give you different options for compressing the video levels.
IMO, properly converting the patterns to the DVD-Video format is probably the most reliable way to go. If you simply display them as still files (JPEG, etc.) from a flash drive or disc, then the results may be less predictable, and you'll need to pay closer attention to black level settings (and possibly other image display settings) on both your player and TV to ensure that they display at the correct levels.
If you are trying to use them as an surround reference, then the correct output levels are important. If you are just using them for geometry adjustments though, then the difference should be more negligible.