Chad B's custom test patterns

Here's some test patterns I made on my Quantum Data pattern generator. I converted them to .PNG format and double checked the windows against CalMAN RGB tripletts. They are all generated with video levels, with black at 16. If playing on a media server, RGB PC levels should be used. Also, the .PNG format makes them ideal for use with the VideoForge pattern generator, which is how I use them now.

First are some normal (non APL) very small windows. They are approximately 4% size. normal 4% windows with pluge.zip 116k .zip file

Second are the same window sizes but in APL flavor loosely based on inspiration I got from the AVS 709 APL windows- thanks Alluringreality! They are in between the AVS 709 large and small APL size. apl 4% windows with pluge.zip 523k .zip file

Third is some miscellaneous patterns. custom test patterns.zip 97k .zip file

The gamma pattern is especially useful for checking 10 point white balance control mistracking. Example: with settings at default, take the blue (or whatever color you want to check) control at, say, 80% and turn it all the way down. Then display the gamma pattern. Run the contrast control up and down and observe if the discolored square remains at 80% or if it moves up to, say, 90%.

The color pattern can be used with the color filters for a quick meterless check of color decoder nonlinearities, as it has matching levels at 50, 75, and 100%. If red, for example, blends in with the gray surround at 50% (bottom horizontal row) and 75% (middle row) but is darker than the surround at 100% (top row), you know your color decoding is nonlinear and you can take steps to correct it (lower contrast, change panel brightness, etc).

The aspect pattern is handy for adjusting anamorphic lenses or lens memories on front PJs.

NEW!

AVCHD ISO image thanks to Vega509

75% Saturation / 75% Level APL Gamut patterns
aplReducedSaturation.zip 105k .zip file

75% Saturation / 75% Level normal approx 4% windows
standardReducedSat.zip 21k .zip file
Edited by Chad B - 3/14/13 at 10:51am

For some reason the Quantum Data software does not have a 15% option, so I used 13%. I could fix it with other software, but it still gets the job done just fine.

All patterns are 100% saturation/75% levels.

Thanks, glad to do it. Yeah, that pattern makes it super easy to identify mistracking and is probably the most useful pattern of the bunch with all the new Sharps, Samsungs, Pannys, JVCs, etc that have 10 point controls.

Well, that's still a bit undecided. I will say I have used normal ones a lot more and gotten great results, but the theory behind APL patterns makes sense to me. I can't give you a clear answer on that.

Sounds like a levels mismatch. The patterns are made to video levels, with 16 being black and 235 being white. The WDTV should be set to RGB high PC levels for them to be right.

BTW, I shouldn't have said the window patterns have a pluge; in order to be a true pluge they should have a 2-4% bar, which they don't. But they do have the BTB bars, which are the other part of a pluge pattern. The brightness pattern is a true low APL pluge.

I just did some exhaustive testing on my GT50 and 5% normal windows give great results taking many things into account; better than my APL windows. The 4% normal windows I posted above should be close enough to 5% to give nearly identical calibration results.

The brightness pattern top row numbers indicate what video percentage each bar is. The bottom row is just a mirror image of the top row. Use it to observe shadow detail performance and set brightness.

Absolute best black level method (my preferred method): Lower room light and get up very close to the display (it's OK to even cup your hands around your eyes if room light can't be extinguished and get within a few inches of the screen on flat panels, and it's OK to look slightly off axis at LCDs). Set brightness by bringing brightness up to the point where, standing up close to the screen, you can just barely make out the outlines of the BTB bars. Then back off brightness click by click until you can't make them out at all. In a nutshell, you do not want the black background to illuminate or dither at all above the MLL of the display.

More shadow detail method (OK for day modes or if you value shadow detail over black levels): do a similar procedure but in normal room lighting and from the normal viewing position. Since it is done from the viewing position in normal lighting, chances are the black background will illuminate just a bit in absolute terms.

Also use the brightness pattern to observe coloring and distinction of shadow detail. Make sure the bars don't have an obvious color tint and/or abrupt color shift, and evaluate how visible the shadow detail is.

Instruction for contrast pattern coming soon.

Contrast pattern:

There is a pluge, starting at left with -2,2,-4,and 3% gray bars. That can be used for a quick visual check of brightness. There is a cross indicator at the middle to aid with meter positioning.

CRT instructions: watch for blooming (fuzzying and expansion) of the middle white block in relation to the medium brightness blocks above and below. Adjust contrast to a point just below where this occurs.

Normal instructions: There is a 104% strip in the middle of the white block and 98% strips surrounding the 104%. Ideally the contrast should be adjusted to a point just below where the 104% strip begins to discolor or blend in with the white block. You want to see the 104% strip and it should not be discolored.

No WTW instructions: Adjust contrast to a point just below where the 98% strips blend in with the white block. You want to be able to see them.