Color Correction Algorithm or Correlated Color Alignment or … oh I can’t be sure so I will give up on the acronym game. I do know what it does and how it works though … or rather how it is supposed to be done.
This system was developed by Texas Instruments (TI) for their DLP chip technology so it is not found on LCD or LCOS projectors. It is strictly for DLP projectors that use color wheels (single chip) so the three chip DLP units don’t apply here either. TI realized since the beginning, that the many single chip DLP projectors out there in the market all had to use a color wheel to create color in this technology.
The color wheel uses color filters, typically red, green and blue to create all the colors in the image. Some of you may have also heard of 5 segment or 6 segment color wheels that also added additional colors like dark green or cyan or yellow and so forth. A display system that didn’t just have three colors, but yellow too … and not only yellow, but cyan … hmmm, this sounds suspiciously familiar. Sharp Quattrons anyone? The addition of more colors into a TV system beyond just the traditional RGB did not start with those Sharp displays at all. The DLP technology had been doing it years before Sharp brought this to market for their flat panels.
Now, although specifications for color filters on those color wheels exist, the color filters may still end up varying some depending on the manufacturer/supplier end. All we have to do is take a look at the blue filters that are typically used to adjust color and tint and found in every commercial test disc on the market from DVE to Spears & Munsil to the Disney WOW packages and of course the venerable THX calibration/demo discs. Almost like NTSC … Never Twice the Same Color.
Sorry, the world is not quite perfect. Everyone follows the same specifications and the filters still have a variance.
The CCA system seeks to account for this by offering a system that is essentially divided into two parts.
- Measured Data
- Desired/Target Data
The measured data portion of the system requires that the CCA system be disabled so that the native colors of the filters are displayed without any color correction going on. At this point, one would take out their spectrometer like the Jeti 1211 or even something like the I1 Pro and measure the various colors on the filters. Typically, one would take readings of the red, green, blue, and lastly white and write down the specific x and y values along with the Y luminance value for each color. If there are filters for other colors, best write down those coordinates as well to be safe.
This collected x and y data is then manually entered into the Measured Data section of the CCA system. You are essentially telling the projector where things really are starting from in terms of the color filters on the color wheel. This is like being able to give a friend of yours directions to drive to New York City when he finally tells you he is actually in Denver. (Without a starting point, you can’t really tell him anything.) Without this information, the projector assumes that all the filters are exactly to their proper specifications which could be for a Rec. 709 gamut or a DCI gamut or something else.
(As far as test patterns are concerned, window or full frame patterns should work the same here. A full frame pattern may create more light output that causes more reflected room light back onto the screen, thus altering the color of the test pattern is a greater way than a window pattern. This is where the color of the room can affect the calibration. A red room will result in a red light reflection back onto the screen akin to a man with a red flashlight sitting in the room. The reflection will affect the image and how the instrumentation reads the test patterns.)
Math wise, this is what it looks like in terms of data:
- x = A, y = B, Y = C to start out
- x = A-5, y = B+9, Y = C-10 (from actual measured readings)
A correction factor can now be determined. It is sort of like triangulation.
Since the CCA system now knows where it is really starting from, it can calculate the correction needed to take you to where you want to go. You can now put in the desired x, y coordinates for whatever color gamut you are working to achieve like Rec. 709 or Rec. 601 or something else. (Rec 709 in the case of THX) This goes into the Target Coordinates area of the CCA system. Desired or Target or Destination is the typical nomenclature for this stuff. Turn the CCA system back on and all the colors are now correct. (Kinda sorta)
Now, of course, there are potential issues with the CCA systems since this is not a perfect world and some of the projectors out there do bad math from what I have seen. (The CCA in BenQ projectors need a lot of finessing while the systems in the more upscale projectors are almost always spot on.) Try figuring that out. So when I buy a $10 calculator, it might give me bad math compared to a $100 calculator?
Well regardless of who is to blame, often enough, you will find that when you enter in the coordinates for white at x=0.313, y=0.329 and then you turn the CCA correction back on, white does not measure 313,329 at all. It may still be a ways off. So now, you fudge it.
You now will take readings of RGBW and manually add or take away x,y values until white actually measures correctly from the instrument. Depending on the DLP system that you are working with, a Y option may be available as well and you would add or subtract numbers here to get the intensity of each of the colors to be correct.
In the end, you still need to verify all your changes by watching known reference material as good looking graphs are no guarantee that the images look correct.