One of the (very few) drawbacks of OLED technology is image retention (IR). If an static image is left on the screen for an extended period of time, you may still see the image for a few seconds after the scene changes. The best example is a station logo (CNN, ESPN, etc) after watching the same channel for hours then switching to another channel. This is not a big deal as it does go away, but can IR impact readings and measurements during calibration? This question is really important now that we are calibrating HDR as panels and test patterns are brighter than ever!
In this article our friends at Rtings.com go deep into IR and it's evil cousin burn-in.
I started to investigate this after running into a peculiar problem...after calibrating the color gamut with the TV's CMS controls, each color would have slightly different x,y,Y values after the final sweep of the TV. Turns out the culprit was IR. During calibration a test pattern may be on the screen for several minutes as the calibrator is trying to dial in a specific color. If I am calibrating yellow and switch to red, the yellow test pattern may stick around for a while and pollute the red test pattern.
Luckily CalMAN has a very helpful feature called "Full Field Pattern Insertion". As the meter is reading continuously, FFPI will trigger the generator to put a pattern on the screen for a specific amount of time to allow the panel to "rest". The key is to trigger the pattern before the panel's ABL (automatic brightness limiter) kicks in. The pattern should stay on the screen for a long enough time to let the pixels to "rest". These times are variable based on the TV's settings and the age of the panel, so how you set it can vary from panel to panel.
I use FFPI for every OLED that I calibrate, but I do have one more trick up my sleeve. By using FFPI along with this trick it is guaranteed that IR from one test pattern will not pollute the next pattern. There is a test pattern in the Murideo Six G called a "moving ramp". It is basically a grayscale ramp that continuously scrolls across the screen. After I am one calibrating one color I run the moving ramp pattern for about ten seconds before moving on to the next color. After experimenting with this on my own TV and several others, I am finding that the x,y,Y values are the same when I take the final post calibration sweep of the TV.
I have repeated this several times with good results. If you have any other insight or tips/tricks please comment! Thanks for reading!
Please sign in to leave a comment.