Originally Posted by bytor102
1) There seems to be numerous "duplicated" patterns included- e.g. "20% grayscale window" shows up under CalMAN, ChromaPure, etc... Are these patterns all identical, or are there differences between them? If they are the same, are they just duplicated simply to better serve the specific calibration software packages in some way?
Any pattern with an identical name uses the same encoded levels. Basically the only difference is the order of the patterns to support the preference used by the calibration software available at the time the disc was put together.
2) I understand the idea that "windows" should be used instead of "fields" for measuring plasma/CRT devices to avoid automatic "level limiters" kicking in, but is there any reason why "windows" shouldn't be used for other display technologies as well (e.g. LCD)? Why wouldn't a "best practice" be simply to always use windows vs. fields?
There are a lot of possible variables when it comes to measuring a video pattern and trying to make sense of the measured information. One possible variable is that different display technologies, models, and settings don't all react similarly when displaying identical patterns. Since most people in similar discussions are interested in relative information, such as comparing Y measurements for something like gamma, I'm generally in favor of patterns similar to the APL measurements from this project, since similar patterns remove any change in APL or Average Relative Luminance during the measurement series. Generally the calibration industry uses patterns that vary APL and/or Average Relative Luminance during the measurement series, such as windows and fields. Since different displays or settings may not react similarly to changes in APL or Average Relative Luminance, personally I don't see how it's generally beneficial to include those variables in the series of measurement patterns. My personal opinion is that people should probably gather a lot of information to determine how their display operates when attempting to "calibrate", but generally people just want to base a decision around a single measurement run, regardless of how various variables might affect the resulting information.
I think there can be situations where fields may limit errors that would be present when using windows. On my prior Sony SXRD the variable backlight setting tended to produce rather unreliable relative Y measurements when using windows, and I believe relative Y from field measurements were generally more in line with most on-screen images. When the variable backlight setting was turned off then most any measurement pattern returned identical information. Personally I haven't looked at enough various LCDs to comment on how things like variable backlights might affect measurements when using the field or window patterns from this project, but I know my current Sony LCD has a non-defeatable variable backlight. I cannot say if either windows or fields provides any preference on my LCD, I just know that both patterns cause my TV to vary the backlight during the measurement series.