MacBook Pro Retina display burn-in?

For what it's worth, I wrote to Anand Lal Shimpi at Anandtech asking if he had any plans to investigate this issue and post an article on his site about it. I think Anand is in a unique position to objectively evaluate this problem and I hope he intends to do so.

bjiibj

The thing is, LED backlight displays with local dimming still have only a "few" rows of LEDs, it's not like there is one LED per pixel, and since the image retention is clearly on a pixel-by-pixel basis (those of us who have image retention rMBP know this all too well), I don't think it's possible that it's due to the backlight LEDs having different light outputs due to some kind of memory. If that were the case, whole regions of the screen would show the effect, not highly detailed images (you can see in photos that demonstrate the problem, lettering left over from a previously displayed browser window is still visible in image retention; LED backlights cannot cause this as there is not one LED per pixel - from what I've read, it's more like hundreds of LEDs at the most, when of course the number of pixels on an rMBP is in the millions, and the LED "resolution" required to show text would have to be much higher than is actually present).

Not sure why the other party was adament with the back light, but your statement is very accurate. Thanks.

bjiibj

For what it's worth, I wrote to Anand Lal Shimpi at Anandtech asking if he had any plans to investigate this issue and post an article on his site about it. I think Anand is in a unique position to objectively evaluate this problem and I hope he intends to do so.

It would be nice if there was some follow up on this topic from other well respected tech sites, probably would benefit many readers.

I'm back from the Apple Store in Munich and what I saw there made me very uneasy. They had seven Retinas on display. Since the store was packed I only had the chance to test four. All four had LG screens, all four clearly showed signs of image persistence after a mere two minutes. (I just displayed a Safari window and set the desktop background to dark grey. Then I googled for the right Terminal command to determine the display maker, entered it and hid the Safari window when I was done with that. That was always enough for the problem to show up.) The image persistence seemed very uniform to me, i.e. all machines seemed to be affected in the same way,

Just to make sure I wasn't seeing ghosts I grabbed an Apple employee and asked him whether this was normal. He was able to confirm the issue and just said that it's always possible that machines are defect. I really didn't want to pressure the poor guy any further, he seemed in over his head after this exchange.

Next up was my appointment. The only possibility for me, they said, is to let them repair it (I eventually decided to do that) or to send it in. Both people I talked to, some guy and his manager, were aware of the issue (not through Apple but through colleagues), the guy I talked to first even exclaimed "Holy crap!" as soon as he saw me screen (and, seemingly excitedly, asked around among his colleagues whether they had already seen that problem or were aware of it). There was never any doubt that they would repair it (i.e. they didn't try for one second to talk down the issue). But no replacement, only repair.

It helped that I was very prepared, I guess. During the ten minute wait I opened the laptop and a Safari window, so I was able to immediately show off the issue. I also showed him photographs I made earlier of the problem.

I will see what happens next. The manager told me that they repair devices to fix the issue and that they will certainly test for persistence. I will also be able to check everything in person with a technician before I get the machine back. But it seemed to me that they are basically limited to ordering the spare part (a replacement screen - the whole display assembly including the shell - costs 553 Euro, by the way, the works costs 29 Euro, but I obviously don't have to pay for it) and hoping that it is good.

At least this once I will let them repair, but my patience is very limited. According to German law I have the right to withdraw from the contract of purchase if the seller is unable to fix a defect (i.e. I give the machine back and get the money back). I would pain me to do that (the Retina is an awesome machine except for this issue) but I don't have the time and nerves to deal with this crap.

johns1 and bjiibj,

Both seem to suffer from a crisis of double standard. When you say that the array is not pixel by pixel (something pretty obvious) and because of this arangement it would be impossible for this system to be responsible for the IR you are also affirming that the WRONG idea sold here by johns1 cannot be resposible for this phenomenon either, since it also works by AREAS and not pixel by pixel. So lets copy and paste again the document that johns1 has used to justify his idea:

"Generally (Twisted Nematic- TN type) LCDs have a parallel electrical field, so all of the display area can be symmetrically controlled. By comparison, IPS LCDs have asymmetrical electrical fields in SOME SMALL AREAS, the image persistence phenomenon will occur at the asymmetrical electric fields. The image will dissipate DURING POWER OFF OR BY AN IMAGE CHANGE in a short amount of time. This phenomenon is a natural characteristic of an IPS LCD."

So, if the Backlit is not responsible and johns1 idea about the problem cannot be responsible as well based on your argument bjiibj, so what is?? Can't you see how clearly you are getting into confusion, because you simply don't know how this works.

johns1, since you are just taking advantage that I haven't replied till now to assume you can claim out of nowhere that you are correct, can you please explain to me what you haven't till now, how can the pixels stay ACTIVE in according to your explanation based on the document that you quoted eventhough the computer is turned off, images are being changed all the time and the monitor is being refreshed? Please enlighten me....

Now, about the Backlit, the only source of lighting for the pixels in a LED backlit LCD is the BACKLIT!! So how can a pixel be lit without the backlit?

The array that you document is talking about when you keep walking in circles is the array of the backlit dynamic LEDs in located in the display's borders.

This is how it works, the areas are lit based on two line sources of energy that excite the phosphor, the amount is variable and is based on the refresh frequency of the display as much as the PWM. If in one cycle the group of LEDs is lit it is possible that in the other cycle it won't, this is also due to temporal dithering, it is a way of achieving the 256 shades of gray. So, some pixels in this group may not decay in time for the next cycle, remaining active, leaving a few pixels active and keeping them form refreshing and this may cause the ghosting.

johns1, you don't understand why I was talking about the backlit because you don't understand the subject well anyway.

Why don't you tell me how the LED and the pixels according to your.... ""theory"" are lit and kept active, what exactly engnites them??

I've been through two LG displays and finally received a Samsung panel today. I have no scientific way to "prove" this, but I believe my Samsung display is significantly less yellow than the LG panels I previous used. White actually looks white to me. FWIW.

While i think the prooved clip in youtube by a user could be falsely prooved by a camera trick cause of the glossy screen by displaying a image of a exterior device on the display, but if true, that IR is there also after restart, it could also be a hardware defect but also a false programed chip. Since battery and condensers are also memory. Eitherway the battery could wrongly still supply energy to the display but also a condenser not fully discharged. What i would like to know is, if a test could be made again by the affected defice without restarting 15 minutes if IR is still present...

michael.ka Germany

I'm back from the Apple Store in Munich and what I saw there made me very uneasy. They had seven Retinas on display. Since the store was packed I only had the chance to test four. All four had LG screens, all four clearly showed signs of image persistence after a mere two minutes. (I just displayed a Safari window and set the desktop background to dark grey. Then I googled for the right Terminal command to determine the display maker, entered it and hid the Safari window when I was done with that. That was always enough for the problem to show up.) The image persistence seemed very uniform to me, i.e. all machines seemed to be affected in the same way,

This makes me suspect there were bad batches of the LG panel particually from some Eurpean sites, because like other people that have posted saying their LG panels were acceptable, my LG panel takes significantly longer to show faint image retention. Whats weird is that you can select the darkest grey background and it seems to show it more easily as a test then say a similar shaded image used as a background which I can't get to show IR period. No yellow tint or stuck/dead pixels on my MBPr BTW.

I will see what happens next. The manager told me that they repair devices to fix the issue and that they will certainly test for persistence. I will also be able to check everything in person with a technician before I get the machine back. But it seemed to me that they are basically limited to ordering the spare part (a replacement screen - the whole display assembly including the shell - costs 553 Euro, by the way, the works costs 29 Euro, but I obviously don't have to pay for it) and hoping that it is good.

At least this once I will let them repair, but my patience is very limited. According to German law I have the right to withdraw from the contract of purchase if the seller is unable to fix a defect (i.e. I give the machine back and get the money back). I would pain me to do that (the Retina is an awesome machine except for this issue) but I don't have the time and nerves to deal with this crap.

Thats interesting about German law right to withdraw. Hopefully the panel replacment resolves the issue for you. These MBPr's are awesome computers to utilize. I thought the difference in 2.3 to 2.6 Ghz CPU upgrade was worth it as that small difference in speed seems to negate the scrolling pause that occasionally happens when viewing certain web sites. Best of luck to you on your repair. 🙂

Hello,

Here's a quick update on my situation...

After dropping my rMBP off at an authorized Apple dealer this morning for a screen replacement (my only option due to the fact I've had it for well over 14 days now), they called me this afternoon and told me that they weren't able to fix it right now and to come by and pick it up.

Now, this sounds bad, but I'm actually starting to get a bit encouraged. Why? Well, the technician finally confirmed that Apple is aware of this problem and, basically, they just don't have a solution yet, so I should hold off replacing my screen, at least for now, until they have a better idea of what's going on.

First, they're hoping that the issue can be resolved with software, as this would be ideal. I agree, but I find it unlikely. A firmware solution would be more probable if all LG screens were experiencing this issue, but some aren't, so it's unlikely to be a code problem. I'm pretty certain that it's a hardware problem or a manufacturing defect.

Second, although it's easy to see, via a terminal command, which manufacturer's LCD panel is installed in a particular screen assembly once it's part of a working rMBP, there's no easy way to tell when they're just sitting on a shelf in a warehouse with some non-descript replacement part number stamped on them. As such, they can't easily separate the screen assemblies with Samsung LCD panels installed from those with LG LCD panels installed. So, as with the multiple laptop replacements that many of you have been experiencing, it's completely unpredictable. They could replace my screen assembly and it might be Samsung, but it could just as easily be LG, with the problem still there.

So, although this problem is annoying, and totally unacceptable, as long as I know that Apple is working on it, I'm willing to cut them some slack, at least for a few more weeks, to sort this out. If it turns out, as I think it will, to be a hardware issue, I personally would rather just replace the screen assembly rather than wait for a new laptop to arrive, one that might have other defects that my current rMBP doesn't have, and then go through the pain of porting all of my data over to the new one.

Here's what I suggest though...

For the time being, continue discussing and reporting this issue and keeping stats on defective units (i.e. number of bad/good LG/Samsung LCD panels). Keep this thread alive and growing, and keep contacting Apple tech support to complain and maintain visibility for this issue. However, I think we should hold off bringing attention to this in the wider media (i.e. ZDNet, CNN, etc). I'd like to give Apple the benefit of the doubt...at least for a while longer.

Hopefully, this will be fixed soon and we can finally have the defect-free machines that we paid for.

Cheers,

Peter

^ Thats good to hear.

I was reading another online thread http://forums.macrumors.com/showthread.php?t=1422669 which has a poll that interesting enough shows a small percentage of Samsung panels with IR issues also?

The post #15 is similar to what I observed:

Bit of an update that I think is very relevant. After doing this test and getting the IR result I re-read what the original threadmaker posted about the dark grey background being picked.

So I ran it again except this time on the LIGHT grey background. After 25 minutes I can detect absolutely no IR at all.

This test only seems to work on that particular background in "such and such" a circumstance. Basically I'm saying is that IR is there but its so specific a process to make it show up that in normal life you would NEVER see it.

I've been totally happy with my flawless screen until running this test. BUt now that I realise more clearly that this test is so manufactured as to not represent real life conditions I'm starting to think I should just keep it.

BUt depends....on the fence about it.

With the dark grey background I was able to see some IR faintly after 15 to 20 mins, but try a slightly lighter grey background and no IR whatsoever.

High-Death wrote:

johns1 and bjiibj,

Both seem to suffer from a crisis of double standard. When you say that the array is not pixel by pixel (something pretty obvious) and because of this arangement it would be impossible for this system to be responsible for the IR you are also affirming that the WRONG idea sold here by johns1 cannot be resposible for this phenomenon either, since it also works by AREAS and not pixel by pixel. So lets copy and paste again the document that johns1 has used to justify his idea:

Well what we know is that the image rentention occurs on a pixel-by-pixel basis, and that the LCD cells themselves are pixel-by-pixel, but the LED backlighting is not. I can't see how any theory that doesn't implicate the cells themselves could be correct.

The residual electric charges in question (glossed over previously with the blanket term "electric field") could be on a cell by cell basis, maybe in the transistor array. Imagine each cell having a tendency (either via the crystals themselves, some property of the structure of the cell, or some property of the transistor array, or some combination of the three) to retain charge when held at a certain voltage, or when the crystal is held in a certain orientation, for a while. Then the effect would be seen on a cell-by-cell basis.

johns1 wrote:

With the dark grey background I was able to see some IR faintly after 15 to 20 mins, but try a slightly lighter grey background and no IR whatsoever.

I also saw image retention with a blue background - the default Windows 7 blue background color. There is definitely a limited range of colors and brightnesses at which the image retention effect is most clearly visible.

johns1 wrote:

^ Thats good to hear.

I was reading another online thread http://forums.macrumors.com/showthread.php?t=1422669 which has a poll that interesting enough shows a small percentage of Samsung panels with IR issues also?

I started that thread over there. The person you mention in post 15 just received their new rMBP Samsung display today.

http://forums.macrumors.com/showthread.php?p=15492782#post15492782

Using the dark gray background and the checkerboard, I can reproduce the IR problem on my rMBP...very clearly...in about 5 minutes.

After about 5 minutes I notice image retention even on the default galaxy background.