kos213 wrote:
Basically, we need to accept that this isn't exactly a thoroughly tested technology, so chances of getting a screen with problems is much higher than otherwise. Those among us who are picky about these things should probably wait a year or so before picking up an IPS screen (or until Dell and Apple use a company other than LG to manufacture their screens)
Jacques LAPORTE wrote:
I would not say "this isn't exactly a thoroughly tested technology" but rather something like "difficult to test when the manufacturing process constraints are high".
Remember the first panels (week 39-42 2009) where not that bad! But when production rate increased....
As I already mentioned earlier here, there are no white LEDs. Instead (on panel of this grade) blue LEDs coated with a yellow emitting phosphor are used. The two wavelengths appearing white to the human eye, when mixed. Same thing with the fluorescent lamps (CCFL) using also phosphor coating.
Panels using this LEDs technology are particularly difficult to build with color uniformity due to operating temperature gradient on the array of LEDs (variation in temperature can be responsible for shifts in the emission spectrum). This is the thermal management issue pointed by Apple in its Patent. This is IMHO the current state of the art in this grade of displays ... (the problem name is 'backlighting' not LG, Samsung or whatever).
This is my conclusion too.
But another added issue is that the emission equilibrium of the LEDs has a relatively long stabilization phase, making the measurement in the production environment a challenge.
"The final production step for LEDs is the optical characterization and the subsequent sorting into so-called BINS"
This sentence maybe gives a decisive clue : very hard to test in the manufacturing process.
This is why IMO very good panels are so expensive.
The phrase is extracted from a very good article, by Dr. Thomas Nägele, –in the LED professional magazine I received late 2008- regarding white LEDs and measurement standards.
http://files.me.com/jacques.laporte/cbj4c4
Take a look, it is probably a good part of the story
kos213 wrote:
That's a really interesting article!
So what I gather is if the manufacturer could somehow allow enough time for the LEDs to stabilize (i.e., complete its burn-in phase) and then take measurements, discarding and replacing those that are emitting a color temperature that is totally off, we would have more uniform screens. This would, of course, take forever and be completely impractical in a large-scale production setting.
However, my question is, is the yellow (green, pink, whatever) tinge not related to IPS screens? Or have other LED backlit screens been afflicted by the same problem?
Jacques LAPORTE wrote:
IPS tech is relatively old (end of '90s) and has to do with the improvement of the viewing angle of TN LCD panels. The common defect associated with TFT was, at that time, a slow response time (in all the technical literature I have read) but not color shift, as far as I know.
It's really difficult to diagnose such a device only from the outside. That's why I will stop here my investigations and enjoy my iMac.
But one thing I know (I posted earlier articles on that point), the color shift I see on my screen is from blue (top) to yellow pinkish (bottom-right). If I push the white point up to 6750 Kelvin, the screen looks globally less warm and there is no more pinkish bottom corner.
But it's more icy blue, on top. No free lunch.
I have changed the blue-yellow balance but the gradient remains.
It signs IMO a white LEDs thermal problem.
Note : High grade tricolor LED (RGB) backlit panels do exist and are not plagued by this issue. They use special hardware to monitor variability in brightness and color and ensure a low Delta-E difference across the screen. The price is also high grade.
.
