Originally Posted by TomHuffman
Originally Posted by ConnecTEDDD
A lot of Problems are reported that are not visible by using a dE Chart as a verification of performace tool....
There are, of course, many areas of image performance that are not measured by dE. Contrast ratio is one obvious example. Bit-depth, resolution, proper geometry, freedom from artifacts, and freedom from clipping at the top and bottom. There are probably more I am not thinking of right now.
However, there is no aspect of COLOR accuracy that I am aware of that is not directly measurable by dE. LUTs are all about achieving a desired result in color performance. They are not designed to address the other areas of image performance I listed above.
So, some areas of image performance are not measured by dE, but they are unaffected by LUT calibration; and, other areas of image performance are addressed by LUT calibration, but they are directly measured by dE.
When I was running tons of tests with a Temp (Free+Full) LightSpace Licence for a lot of days (over a month) before to decide if the LightSpace software can make a difference that i can see to my setup, (because the money were a lot for a HT user), I had 1 memory with 5-Point Cube, 1 memory with 10-Point Cube, and 1 memory with 17-Point Cube to compare, These 3 memories were generated from the same TV-Settings.
Measuring these 3 memories with 21-Step Grayscale and 5-Step Saturation, the differencies between these 3 memories were in 0.05-0.15 dE2000 Average, but while i was switching them during moving or static known for me movie content, the difference was very noticable.
5-Step Saturation Runs have very few color points for someone to have a total overview of the performance of a display, skin tone colors and other memory colors are not measured with a 5-Step Saturation...
Color Checker 24 Patches are not enough also....Just with Color Checker you are getting a more accurate look of some common colors from the total of 10.648.000 Rec709 VideoRange Colors...
A skin tone has a gradient of Brown/Yellow shade, if you use a large cube like 17p these brown/yellow scale steps are calibrated too.... They are not measured or calibrated with a 1D Calibration or a small cube size LUT...
With a 17P Cube Profiling the software takes 4,913 hard accurate reads, then the software using some 'secret sause 'maths extrapolates those values to 274,625 precision color points … where eeColor uses those 274,625 points accurately reproduce 10.8 million colors.
If you measure that VT60 that was calibrated with only internal controls by taking 4913 of that display (with a Klein it takes 1 hour max.) you will discover a lot of things that are not so good / not visible at classic 2D Software Charts.
Also some patterns from my upcoming disk will be a real changelle for any display that is not LUTed