Originally Posted by Pannus
Thanks for your response. It sounds as if you also have struggled with the Sharp Elite CMS. Set at low temp in THX (which was closest to 6500K), the green and blue primaries were INSIDE the CIE triangle. I am using my lumagen mini as a pattern generator and adjusting the CMS in the set with calman. Try as I might (and my efforts have been HOURS) I cannot get these primaries spot on despite the three controls. The cyan and magneta secondaries are also a nightmare.
You would think the initial primaries would be outside the triangle for this display (the red is). As I just acquired the lumagen, is it possible that the default saturation is at 75% or something?
I hope you come back to answer this!
If any color starts out inside the CIE HDTV triangle, the Lumagen processors (or any other external processor) cannot move the point back out where it belongs. The video display's CMS controls MIGHT move the color farther out. When a color is inside the triangle, it is under-saturated. When using a video processor, you need to find a mode on the display that either places the colors in accurate locations OR moves the color(s) OUTSIDE the CIE triangle. If the color starts OUTSIDE the CIE triangle, the video processor can move it in to be accurate... but video processors can't move colors that are already inside the triangle so that they are farther out.
You want to use 75% color patterns because those are 50% stimulus (roughly, with gamma in the 2.2-2.3 range, 75% patterns should be pretty close to 50% stimulus). The reason for that is... if you calibrate with 100% patterns you are all the way out at one end of the luminance range. A lot of errors could creep in as you move down the luminance range. When you use 75% patterns, you are right in the middle of the luminance range (50% stimulus) so when you make 50% stimulus accurate, you have a good chance of minimizing errors since the luminance range will only vary +/- 50% from where you measured... versus a range of +0% to -100% if you were to use 100% color patterns.
The calibration software generally will have a setting where you can select 75% or 100% patterns for color measurements (and possibly some other levels also). If you want to get picky, you could measure colors at 25%, 50%, 75% and 100% then determine, in the aggregate, what settings for CMS controls produce the overall lowest errors for each of those pattern levels. But you'd be looking at hours and hours of measurements and comparisons.