Originally Posted by ptchristensen
In the service menus for my 34XBR960 I have a menu called QM. This menu contains a massive amount of geometric patterns and color patterns.
pt, you're an absolute Prince for discovering this!
It doesn't appear in any service manual I have seen. Enter service mode, and step *backwards* with the 5 key to see these groups. My 36XS955 has the service-menu group called "QM." It comes up in blue, not green, and AFAIK the values are never saved. Further, 0 - Enter reloads everything and kills the patterns, rather that your having to run the PATN count down to 0 before exiting. So does power-off.
#0 - INFO (values 0-7) has 7 screens of information about what is going on in the set. Nerds, rejoice!
#1 - PATN (values 0-80) has four groups of 12 patterns, both color and b/w, including a couple of amazing frequency sweeps. Each pattern is at value 20*n + p, where n is 0-3 for the group, and p is 1-12 for my pattern set. (The remaining values are pure-white screens.)
The groups on my set appear to be 1080i, 480i, 480p, and 720p. That last one may be exclusively component video, and the middle two NTCS color by a different pathway.
#2 - GPTN (values 0-12) are 11 different patterns of a decidedly different type on only the upper 70%of my screen, no matter the scan rate. Some are translucent. I don't know their specific purpose.
QM has more codes, but I haven't discovered anything about them yet.
The test patterns are inserted into the video channel *ahead* of all picture adjustments in 2170P-3 and MID5, so they can be used for evaluating these settings. I value the crosshatch patterns, but some of the others I will probably just look at for fun, as we don't know what pathway they take thru the long processing chain. I wonder what the sweep frequency extremes are for the #8 patterns in each group. Unfortunately the highest video frequencies are positioned at the right edge of the screen, where the horizintal focus is the worst!
I will bet these are for quick assessment of major probems by a service tech, and that we can trust perhaps only the crosshatch patterns and grayscale steps for calibration accuracy. For me, these will *not* replace the sophisticated patterns available from DVD via component input, nor memory-stick patterns, for center focus and other screen-center assessments. I don't trust the bright color bars at all! The gray step-scales are likely useful for assessing your grayscale linearity, unless you have a color bias on one side of the screen -- and then other patterns would be much better.
Kudos for the detective work!