The take away Avoid extremely long play of static images and elements (variety) Dont use max bright, less is more And most importantly dont leave your steam deck on for 750 hours straight on a static image

Another tip when downloading games: Switch to Desktop mode and in the power settings, enable a check box that disables the screen after x minutes, so you can download all you need while the screen is off.

Incredible that you've sacrificed your Steam Deck and are going to this level of detail. The legend strikes again.

This is why i love youtube,, someone always do stuff no one others would think of doing😅 Btw: This is actualy usefull information, but i came from a video to show if a ps5 controller would melt on a grinder 😅

Valve should let us set the time to darken the screen while doing downloads.

As a comparison with Wulff Den's experiment on the Switch OLED: -> This isn't burn-in. This is subpixel degradation (and the same goes for the Switch OLED test as far as we know). Pixel degradation occurs as soon as a subpixel is on. The higher the luminosity, the faster the subpixels wears off. -> 750hrs at 1000nits isn't the same as 750hrs at 340nits (the Switch OLED panel). -> Even after a year, for his first video on the subject, Bob Wulff said the Switch had "burned-in". His conclusion, after 2 years, was just that this issue isn't as big as some people think it is. -> In the case of Bob's experiment, he used a static game image at full brightness, making all the pixels on at the same time and at full brightness. The wear of each subpixels depends on each pixels color, but all the pixels degraded at the same time. -> In the case of the Phawx's experiment, he used a black background, making all the other pixels off. Since those pixels were off, they endured no degradation. This resulted in a much more visible result as we could easily notice the degradation of the pixels that endured the stress test compared to those which didn't. If we decide to do this kind of test with a full white picture, all the sub pixels would degrade as fast and in the same way, so we wouldn't be able to notice the pixel degradation as it would be homogeneous, and we would need to put our test unit aside a brand new one to notice its lower luminosity caused by the pixel degradation. My point here is that even without actual details on each panel relative quality, if Bob's test seem to show less degradation than Phawx's, it's only because we only look at the difference of luminosity between the pixels that were in the "white zones" of the test picture compared to those in the "darker" zones, but all the pixels on the Switch's panel have degraded. In other words, on the Switch's test, we just focus on the difference of degradation between the different zones, (which degraded for the same amount of time, at the same luminosity level) but not the actual degradation of the panel itself. For that, we would need to measure each pixel zone's maximum luminosity output compared to a brand new panel. In short: those tests aren't comparable. If we consider Bob Wulff's test to be somewhat unrealistic as no one would leave their device on showing the exact same picture (even less at full brighness) for 2 years straight, what the Phawx did is even worse. It was specifically made to see the actual pixel degradation between full-on HDR at 1000nits compared to zero pixel degradation, which make it very contrasty and very clear. While in the case of the Switch test, it's much harder to see the actual degradation considering that all pixels have degraded at the same speed. And no, the somewhat darker zones of the image are not a good reference to consider the full pixel degradation the whole panel endured. In very few words: it's not because you can clearly see that some zones of the Switch test have "burned-in" that the other zones are "as new". The whole panel degraded. You are just looking at a zone that is bit more degraded than the other. This isn't comparable with the Deck test, where the screen had most of its pixels turned off which show no degradation at all as a reference.

I think we have some confusion on OLED burn-in terminology. Image retention isn't permanent, on OLEDs it happens quickly and is fixed by refiguring the voltages applied per pixel (often called a short pixel clean). OLED burn-in (probably better called "burn out") is permanent, and can only be compensated for by wear leveling the rest of the panel (often called a long pixel clean). The image doesn't need to be visible with the display off to be considered burn in on an OLED. Wear leveling will decrease the peak brightness, and eventually burn-in will become severe enough that no compensation will help.

The video i've been waiting for!

Woah! I am impressed by the transition of the thumbnail into the video!

Thanks for advice mate. Your channel is great as always

I appreciate your dedication towards tech while most people would never do this glad someone answered the call and made the sacrifice to enable all of us to be more informed. Glad to see it would take a crap load of abuse to actually cause issues to the OLED panel for the deck it's good to know for future buyers.

Super informative! I also have appreciated the advices to reduce burn in, thanks

Do you have any plans to put the the smach z videos back up? Its a interesting piece of of handheld history even though they ended up dead.

Great video! IMO there will still be a slight risk with black bars due to the aspect ratio, but if gamescope gets an option similar to youtube's ambient mode, it will go a long way to mitigating that.

no one is gonna talk about how clean the transition from thumbnail to video?

Thanks for doing the dirty work Phawx! And for sacrificing a Steam Deck for SCIENCE! I agree with some other commenters, this doesn't freak me out as much as show me how not to treat it and how hard it is to genuinely ruin the screen. Especially given the Steam Deck OLED's handheld / battery-centric nature.

first time watching a video from this channel. gotta say that thumbnail to video transition was fenomenal

Ok how about the battery since its 750+ hours tested? Does it degrade since its on for that long duration.

What's the risk of burn in when just using the fps overlay? I was wondering whether I should turn it off or if I can leave it on.

Now my question is. Does it scale linearly, ie 500nits degrades twice as slow as 1000, or is it slower. Comparison between 1000nits at one point and then looking when 500 gets to similar damage

Does the Steam Deck have any "pixel refresher" function like OLED TVs and monitors? Not all image retention issues are permanent on oleds.

Thanks for testing this.

Did you check at any earlier marks than 750 hours? Would be useful to know like how much burn occurs at 100 hours, 200 hours etc so we can gauge how many hours we can put in on a game with static elements before burn occurs

it's impressive what you have done. not only with the testing but with how you make the screenshot become the video.

Good informative video, thank-you. Can you do this expensive experiment again with hdr disabled?😅

I really like how you intro your videos

So I have a little experience with this, I had a LG OLED c2 in 2017 and I had to have the screen replaced 3 times in the first yr of the warranty, I just sold it after the final one. I used to to play games a lot and really the amount of burn in was insane it only took a couple months for severe burn in to happen. and I really tried to not have images in the same spot all the time and consistently refreshed the pixels. So now, 7 years later I got myself a Asus ROG Swift PG48UQ 48" OLED gaming monitor. The difference is insane. I've had this monitor running probably 22 hours a day no exageration and there is not even a hint of burn in anywhere on this monitor. Even small text is nice and crisp (where the old OLED has blurry text) and as long as I religiously refresh the pixels I'm hoping this monitor lasts me years. They clearly have improved the tech because i run windows on this monitor and there are icons along the bottom that are there all day long and there is no burn in. But this is really bad what he's showing, it shows that pixels are dimming over time.

For your future videos covering this subject may I suggest looking up some other panels and tests online so you can present them as a reference? (Switch, TVs, Monitors) Could be helpful for people to know that it's not Just the Steam Deck with an OLED problem. It's OLED, they all have this problem. Probably would be hard to tell if others are worse/better than the Steam Deck panel unless similar testing was done on them. But it would be educational to know where we are headed in the future of OLED tech.

Compared to Wulfden's video its either Steam uses a different panel as the Nintendo Switch Oled, or they cranked the brightness really high (I would assume this one, since IIRC Switch doesn't support HDR)

Awesome vid and detail, I'm not sure why everyone only cares about burn in in OLED monitors tho, my 144hz 1440p aoc VA monitor has the windows search bar permanently burner into the bottom left corner

Does the steam Deck OLED have pixel refresh or panel refresh cycles?

This is why i replaced my OLED with the Legion Go and couldn't be happier. Sure no perfect blacks but everything else is soooooo great

Why do we see the sdr90 ”burned in” when a wallpaper or cyberpunk is displayed but nothing on statics red green blue, even on static grey there is nothing ?!?

Great video and I can see you're very knowledgeable, I look forward to future videos, subscribed!

Just curious, with TVs and monitors you typically have refresh cycles which work to combat drift in the TFT matrix which can cause image retention, rather than retention caused by degredation of the organic material. Is there any form of screen refresh on the steamdeck, or are mobile oled panels different in general from large format displays due to how they're driven?

Interesting test. I wonder if there’s a way to run a compensation cycle on the deck (idk if that would be something run in Linux or would have to be built into the display controller). rtings’ oled tv testing has shown that running a cycle can clear up a surprising amount of image retention

750 is pretty quick, though your test is an absolute worst case scenario. Does the steam deck not have a panel refresh option?

So because this is image retention will it eventually go away? Is there a way to fix it?

I’d love to know if there is a way to run short and long competition cycles on that panel.

For downloading games, you can turn on allow background downloads, launch a browser, and visit a dead pixel test site and set the color to black and fullscreen it. Your OLED screen will effectively be off while things download.

damn that OLED screen looks so much better than my OG 512er. Never really got it that clear before.

I wonder if there's a difference between a sustained image on screen vs. intermittently displaying that image mixed with others. For example, does leaving it on one static image for 1 straight hour affect the screen more vs. intermittently going between that image and others for the same total time (but broken up)? Basically, I'm wondering this because the latter is more of a common use case.

well ,well, well (eats apple while staring at a OG Steam deck) thank god for Vibrant Deck

how did you even activate HDR? I don't find the option anywhere. Also not with dev-mode

Remember kids, nobody leaves their OLED screen showing a static image for 750+ hours.

Now we can get a replacement screen video :) Are the OLED panels available for purchase?

Hang on... I was under the impression that image retention was temporary burn in and Actual burn in doesn't require you to be able to see it when the device is off. Was I wrong? I mean I was given a plasma TV, and when it burnt down I couldn't see it when the TV was turned off. Also After 8 months of owning a LG C1, I had image burned in to my TV for 12 hours. I assumed this was imagine retention I've heard TV reviews talked about. The image was black bars for watching Anime. After that, I'm never buying OLED again. Sure it looks Superior to any TV I've owned or looked at, but the stress of it happening again is too much for me to bear, so I'm saving up for its replacement.

I think this should actually give people a sigh of relief to know that regular or even heavy "standard" use shouldn't cause any issues. This was test was done with the intention of damaging the screen and even then the damage is not so bad that it renders the display useless. Sure- it has image retention, as we expected but tbh if you're using your Steam Deck like this I think this can be expected and would pass as being expected. Now what would be concerning is if the panel shut off or permanately killed any of the pixels in under 1000 hours. But from what I see we're getting exactly what we expected here. I was super curious to see how this panned out ever since you mentioned it to me and I'm thankful you made the sacrifice so others didn't have to! Awesome video, awesome application.

I have like 80 hours of gameplay already on Forza horizon 5 on my 1tb Oled and no burn at all, the 1tb limited edition Oled is a boe panel, the regular 1tb and 512gb Oled are both Samsung screens

I don't know what you wanted to show with the screen off but you won't see burn in with the Steam Deck powered down. Also this is burn in and not Image retention. And it's a really big problem for most OLED displays. Only thing Valve can do is a compensation cycle where all pixel degrade to an even level. But this would kill color vibrancy and might not work with severe burn in.

Watching on my SD OLED. This was very interesting to watch.

Good to know I'm pretty safe here 😂 I play my deck for maybe 2 hours max in a session if I have the time, and I play plenty of different games so static HUD elements aren't on screen too often. Plus I'm never at full brightness cause around 50% is perfect for most environments I play in, and I'm always at low brightness at night which is about 60% of my playtime. It's only ever on a static screen when installing games or downloading things in desktop mode but I always turn the brightness to 0, keeps it from retaining images and is less distracting so I can just let it do its thing Love to see good OLED panels that aren't too susceptable to burn in/image retention besides extreme circumstances: A. Panel aging, having the thing for so long, using it for years and it just getting old B. Really trying to damage it/being careless with the screen. Pretty good stuff. Good to know in my use case I'll be fine for a long time

The thing preventing me from getting a Steam Deck oled is Monster Hunter, the whole game revolves around static images everywhere.

So you were not able to verify if this panel properly runs a pixel refresh cycle, or does the steam deck offer a pixel refresh cycle at all even? Btw you dont see burn-in on OLEDs when theyre off, since they are self emissive pixels and so when there is no current they are completely black. This is different from a CRT where the carhode ray burns-in to the phosphor layer itself. My comment before was more about buggy pixel refresh which may cause image retention to permanently show up as if it was burn in, untill a long pixel refresh is triggered. I suspect that may be the case here. Though having those images be stuck may cause burn in of itself so a good pixel cleanup cycle is essential to OLED longevity

Burn in will very rarely show up when the display is off, I've seen plenty of burned in oLED android devices with very heavy, very clear burn in, way worse than your Deck and none of them showed anything beyond when they were powered on. As a Deck oLED owner myself now, I can appreciate your sacrifice far more than before. Gives an excellent indicator the expected MTBFs of the display, though of course burn in isn't exactly a failure.

A few reminders to folks: 1. OLED pixels wear down with usage - that is a simple fact. 2. This test is a "kill the pixels as fast as possible, and as noticeably as possible". 3. The reason the image retention is so prevalent here is due to the "as noticeably as possible" aspect. 4. Bob Wulf's test and this test are not the same test. To put it in different terms, a shirt is going to have its colour fade as it ages. This test was as if someone sectioned off parts of the shirt and put different chemicals on it to accelerate aging at different rates. If you picture how that shirt would look, and how you would be able to clearly make out the sections that had the chemicals, then you can understand the results of this test. To continue with this analogy, Bob's test was (more or less) unevenly spraying the entire shirt with a single chemical and then watching the shirt age unevenly.

Also u can plug in a monitor to mitigate image retention

Would be interested in seeing burn in comparisons at the same brightness level.

Nice to know. I got my OLED 1 month ago and downloaded 6 hours with 70-75% brightness. Shouldnt caused any "burn in" effects and in the future I turn down the brightness to the minimum. Thank you 👍

This is what I've been thinking when I'm about to get an oled steam deck and I didn't Oled is not for handheld mini led or micro led is the way to go

Are you going to replace the panel once you are done testing?

So, if you put a static image on the screen, in HDR, at max brightness, for over a week straight... You get very slight degradation. I think we can confidently say this will not be a problem under normal use, but that is an argument for making sure the screen is dark first if you're going to plug in your device and then not pick it up for more than a day or two.

I feel that right off the top you should have mentioned you had one static visual on the screen for 700+ hours. Not an app, which can be confused for something with motion or usable elements.

I expect a lot of second hand Steam Decks with 'Meet Hanako at Embers' burned in.

My girlfriend has burn in from the Hud in star dew Valley :( which is the game everyone says to play. And she plays in the dark mostly so get screen isn't even max bright.

Thank god for showing this!! I guess I wont be having to upgrade my LCD, I love my current steam deck. Dont get me wrong, OLED is nice but, I dont want to see any burn in or risk damaging it on the long run. THANK YOU!!!

What I take away from this is, it's not an issue unless you are aiming to make it an issue, and green oled pixels are immortal.

I like how youtube channels have evolved in terms of hardware testing. Back in mid 00's, NOBODY would buy say, an XBOX 360 and leave it on for days running a intensive game to see what happens.

So my thinking is that the color subpixel (red green and blue) are loosing their respective color brightness luminosity. So rather than "Burn in" it would be the subpixel is "burning out" red subpixel cannot be as bright as it used to, because its been at max luminosity for 750+ hours "Burn in" is the retention being baked into the display - even when its off.

Where did you get the shelves behind you?

Can you "unburn" them using a different color in the same place?

I’ve always just connected my OLED deck through a dock to tv while downloading titles since day 1…problem solved. I never use max brightness anyway. I’m always inside if gaming

I was just thinking about this yesterday when I had to leave the deck on to download bg3

does the steam deck oled not have any burn in prevention measures like pixel shift or pixel refresh / compensation cycles etc?

I'm glad, i bought steam deck vs old IPS matrix ❤️

So do you suggest to play game with no HDR?

Can valve add a auto dim/screen sleep for downloading etc already??

Really want to see same test on the limited edition version that uses OLED screens made by BOE. Regular OLED uses Samsung screen.

I've had on oled tv now for 6 years and played no end of 50 + hour games in HDR with differing static elements, and there is absolutely no sign of image retention. I occasionally check out of interest and there is nothing. I think people underestimate just how extreme these tests are. Trust me the screen will outlive the sticks, battery, buttons and just about everything else on the steam deck! By that point steam deck 2 if not 3 will probably out. I would say forget about it and enjoy 👍🏻

Can't someone make a Decky plugin to enable things like pixel shifting?

I think your test is overblown 😂

are you guys still going to buy oled SD? 😅

I run it at 40% brightness.. The eyes adjust and will eye strain at 100%. 40% is like 100% if you aren't out in the sun.

Now you need to do some inversed burn-in😂

Can you just run a noise pattern that would reverse or minimize existing burn-in?

Wow Bob Wulf did this on his Switch OLED and that seemed to last much longer than this......i wonder why the Nintendo panel lasted so much longer?

A lot of people seem to be playing this down as if they don't have static elements on screen, performance overlay anyone? The fact that even the low brightness SDR 90 text it was showing up clear as day after only 750h should worry people.

During the red section I think I noticed flickering by the h in the hdr1000

OLED screens are gonna OLED.

Awesome I was going to buy it thanks I will buy the odin 2 pro better battery life and 1080p screen and a lot smaller

Yeah you burned in the text, but usually you don't get burn in's since you either push the power button or put it down for an hour at most while paused. How long did it take to burn in the screen? Did it take 500 hours or 30 hours or 100 hours, it's a strange scenario though having the screen on for like 72 hours straight on one and the same frame. So there shouldn't be a risk for any burn in's at all.

Why would we expect to see any burn-in image when powered off? Oled only emit light when powered.

You shouldn't be responsable for how people takes your conclusions, it would be very interesting to see how different oled technologies behave, maybe that can put the topic on the spot with Valve looking for some mitigations.

So it's more like the pixels have burnout rather than burn-in. 🙃

People who are harping on about "don't keep static images on and you'll be fine" are missing the point: Burn-in is cumulative. Meaning - the degradation of the pixels happens over time due to usage (think of it like a candle burning; more accurately it would be called burn-out for OLEDs). Dynamic content will help with not seeing it as soon as on this test but it's INEVITABLE. But that's something we already knew from years ago from TVs, and only the most deluded OLED fanboys pretended like it was no longer an issue. The concerning thing here is the time-span. Even the old TV tests RTINGS did for B7 and the like (which more easily burned-in) only saw burn-in appear the earliest at around 3000 hours (CNN max brightness stress test). Considering game HUDs are usually in the same locations, even if you baby the deck and keep it a lower brightness, let's say it doubles the time before you see it, so 1500 hours. That's basically 2 years if you use it 2hrs/day. Not great imo, but not a tragedy either. Still better than putting up with the schitty LCD panel from before. Just know what you're getting yourself into.

so, how to resolve the problem ? If this is not burned in, it's fixable ?

well I won't be playing tekken 8 in HDR then. That's a game that I would likely play for 500 hours with those bright life bars always being on screen.

All I know is my oled steam deck has a nice crisp picture I also never run my device over 50 percent brightness and most times my devices are lower than that I can’t have the screen too bright or it hurts my eyes so I probably won’t see any of these issues

Oh noooo....I just got the oled today 😢😢😢 my original SD didn't burn in at all and now I'm afraid 😳

This gives me more paranoia about keeping the lvl 2 performance enabled while i play.