Originally Posted by Wizziwig
Direct quote from their article
"We contacted LG regarding the strange results in week 4. LG engineers visited our lab a few days ago and were able to confirm the 25% window on the Live CNN and FIFA 18 TVs are a result of a factory issue (see our video here). OLED TVs are produced in a hot process, and after cooling a 25% window is shown on each panel. Some TVs which haven't cooled completely can produce invalid results for the lookup table used by the 'Pixel Refresh' function, causing this 25% window to become visible. Only some 55" OLED TVs were affected during part of 2017.
As this is not an issue with the panel itself, it is possible to apply a fix to the lookup table. LG will apply this fix to anyone who presents this issue to their support, for free, even after the warranty period has long expired. They have fixed our two affected TVs (see the uniformity photos below). Note that this doesn't fix other uniformity issues as the result of static content, only the 25% window caused by a factory defect. LG has also confirmed that there is variation between panels, which is why some OLED appear more prone to developing uniformity issues (as in the case with our Live CNN (200 nits) vs Live CNN (Max).)
If you read the rest of their article, "uniformity" in their context refers to the non-uniformity caused by the burn-in.
I suppose you know more about these TV's than LG engineers
No, I don't. But I also don't take statements to a 3rd party reviewer at face value. I will generally steer towards the simplest explanation for a mystery/problem until I see irrefutable evidence that the explanation must be more complex...
Yes, my electrical engineering degree and over 20 years working on hardware and software in the video game industry mean nothing.
Fair enough - did not mean to offend. Our backgrounds are similar and the gao between how we view these issues seems more related to our approach than technical background.
You prefer to see the glass half empty and problems complicated enough that resolution will be difficult or impossible, while I prefer to see the glass half full and problems that are solvable given enough time, resources, and priority.
If I was responsible for what LG has achieved with WOLED over the past 5+ years, I'd be pretty darned proud. Listening to you, sounds as though if you were responsible, you'd consider yourself a failure...
The kind of predictive compensation algorithm you're dreaming up can't work in the real world unless every panel is an exact clone that responds equally when subjected to the same signals and usage history.
No, that's an exaggeration. Panel uniformity merely needs to be within perceptible limits. My suspicion is that the limits of this approach will be much more limited by ability to accurately accumulate and estimate cumulative usage rather than panel-to-panel variation. But at any rate, the 2017 rtings.com results indisputably prove that LG is applying some type of bur-in compensation...
LG already admitted above that this is not the case.
LG will say anything to buy more time on this issue. I don't believe anything they say on the subjects of burn-in or uniformity. I'm not even sure the 'engineers' they sent to rtings.com truly understand the science of what is going on at the level we might expect...
We've had owners like
who already told you months ago that their two E6 samples which were used for similar content
showed differences in their burn-in resistance.
Nice anecdote but not nearly quantitative enough to be definitive. I mean, come on, who waches the same
content on two TVs
rtings.com has the rig set up to truly display the 'same content' and when they get different results in two identical TVs with identical settings, that'll be definitive and I'll switch camps.
Until then, it's just too much of a cop-out to blame panel variation, throw your hands up in the air and say 'all is hopeless with WOLED and we're all doomed to suffer from a mega long-term panel lottery which will take years to know if we've gotten lucky or struck out...'
You're also conveniently ignoring that 2017 model owners like
developed traditional "dark" burn-in on their 2017 OLED
despite having this new wonderful compensation algorithm. On his unit, the predicted compensation amount was clearly too low.
This is a good point (dark burn-in one one user sample) but there are alternative explanations.
One explanation is that LG focused on CNN burn and put all hands on deck for that specific case at max OLED Light. So that specific cause of burn-in is relatively well-compensated for in 2017 WOLEDs but all other sources of screen burn are incorrectly estimated and compensated (both high and low).
The fact that there have been so many fewer reports of burn-in on 2017 WOLEDs (so far) than we saw for 2016 WOLEDs speaks volumes (and more loudly than any individual anecdotes). Now, it's still early for 2017s, but within the next ~3 months we should be in a position to make an apples-to-apples comparison between 2017 and 2016 WOLEDs as far as susceptibility to CNN and MSNBC burn-in...
On the rtings.com samples the boost was too high. You will never guess the correct boost amount without a way to externally measure the light coming out of the panel or can produce them in such a way that they all age exactly the same as your wear prediction model.
We can both agree on the part in bold. The rest is non-constructive exaggeration.
Are you even reading anything in this thread? Just a few posts up you have owners who reported this 25% rectangle problem to LG back when the 2013 TVs were released. Some even had their panels replaced under warranty for the issue back then. Problem still persists even on the latest models just released.
I looked at a store demo of the 2018 55C8 this past weekend and it still suffers from this issue (along with same poor uniformity). See attached photos. Rtings.com only tests 50% gray backgrounds. I'd be curious if their near-black images also show a "fix" to this calibration rectangle retention.
I havecto admit that in my several-years-now of perusing WOLED owner threads, the rectangle as a significant PQ defect escaped me. Stuck subpixels, vignette, anti-vignette, yellow-piss-stain, off-axis color-shift all appeared much more worrisome and significant than this rectangle that many assumed was heat-related (incorrectly, it seems).
Certainly I hope you'll agree that to the extent LG had to prioritize which fires to put out first, addressing PQ defects like vignette was much more important than addressing whatever causes these feint near-black rectangles.
So again, glass half full versus half empty. I'm impressed that LG is now owning up to this low-priority issue and promising to make affected owners whole (forever) while you feel it is unconscionable that LG has taken so long to own up to this obvious and glaring flaw.
I did go back to check my 4% and it appears I won the panel-lottery on this 25% rectangle issue as well...