12 ms·
MicroLEDs Moving from Lab to Fab
- jbuhbjlnjbn 5y agoVery interesting article. The writing style of author Laura Peters stands out to me, it is informative, filled with many short background explanations suited for laymen, to the point, factual, entertaining. The possible leap in responsiveness cannot be overstated, such displays will probably solve latency issues plaguing alternative display technologies for good. I cannot wait to have 1000+ Hz ultra responsive screens in any imaginable format.
- giuliomagnifico 5y agoYes! I posted it for this, it’s nice to understand how a new (quite) technology is going to mass production.
- beebeepka 5y agoOLED is plenty fast. It's everything else that is holding us back. It took the TV companies decades to move, reluctantly, beyond the silly 60hz cap I am all for 1000fps games but by that time I won't be able to play them due to old age.
- skybrian 5y agoI don't understand why you care? 60hz seems plenty fast enough and there are a lot of things slower than that. We are usually waiting on networks, not displays.
- rbanffy 5y agoGames can push more than 60 FPS easily, but a lot of TVs (I’d assume most) are not used for that and manufacturers can extract extra profit from premium models that enable that (even if the panels and electronics are identical because it’s cheaper to manufacture that way). For me, personally, the benefit of an HDR panel would be much higher than faster frame rates.
- Dylan16807 5y agoWhen you're interactively controlling something, you're best off getting new frames in a small number of milliseconds end to end. Lots of the delays in the pipeline are measured in multiples of frames. There are serious improvements in the ability to click on things up to 200-300Hz, and going above that is less important but still not bad. For something that can toggle as fast as an LED, might as well go for 500/1000 and minimize jitter.
- skybrian 5y agoIf the rendering pipeline adds multiple frames of latency, it seems like that's a separate problem that should be fixed and it would help a lot without raising the frame rate.
- Dylan16807 5y agoYou pretty much have to buffer a couple frames because that's how rendering works. You can do work to reduce extras, but it only gets you so far. And screens tend to want to buffer at least a frame too, and it's hard to convince them not to. You also need quite high framerates to make rendered motion come out reasonably smooth. Especially for something you're trying to follow across the screen with your eye. Fake motion blur only partially helps and has its own problems.
- skybrian 5y agoI wonder if it makes sense to render at 120hz internally but skip every other frame? That should have a similar improvement on latency as actual 120hz. I agree that following things across the screen is a use case for higher rendering rates.
- Dylan16807 5y agoThat helps, as does triple buffering at very high framerates, and various techniques to try to align rendering exactly with when a frame is needed. But it's still significantly better if you put that 120+Hz output onto a 120+Hz screen, and it's not super hard to make them.
- rbanffy 5y agoTVs have longer lifetimes than phone or laptop displays, so it’s safe to assume manufacturers would be more cautious. As for TVs, there’s not enough 60+ Hz content to justify much of a hurry. I’d be surprised if there were a significant amount of content above 24Hz.
- ksec 5y ago>Haynes noted, however, that the timeline for commercialization of large area displays on TFT panels “will depend on how quickly the industry develops mass transfer technologies that can deliver high accuracy and exceptional reliability.” Indeed, Apple’s latest iPad, which uses white miniLEDs for backlighting, requires the accurate placement of 10,000 emitters, which sounds challenging enough. Positioning 25 million red, green, and blue subpixels on a 78-inch TFT panels for a 4K TV, with a tolerance of only 10 non-functioning pixels, will require an exceedingly robust, highly accurate transfer method. >Different Paths to RGB...... Basically uLED Backlight with QD And the rest. As much as I love and want uLED. None of these are new, and the problem has been roughly the same from day 1 in 2016 when all the media were jumping on uLED. Even if they could solve all of that they still dont have a roadmap of cost reduction that could match OLED. On the other hand OLED cost has been dropping quicker than expected. It is another one of those thing like Solar panel where all industry analysis and projection were wrong. We also have a solid roadmap to improve colour accuracy, burn rate, efficiency etc. Along with other "potential" OLED adjacent technology like QNEDs where it is 80-90% of uLED but could be used on existing OLED manufacturing. Again as much as I love to see uLED, objectively speaking I just dont see how it could compete. And it is one of those moment where better is the enemy of the best.
- sschueller 5y agoI am old enough to remember when OLED was in the same boat. Yield was low, expected life was short and RGB was far off. I was extremely excited when I got my hands on a 2 line character single color OLED display many years ago. That a phone would ever have an OLED color display was a far off fantasy. I am sure we will see uLED at some point, it just takes time and a lot of research.
- fsloth 5y agoYes, I remember when OLED:s were far off future technology for mass consumption. I think the first OLED display I saw was a tiny Sony PAL TV that probably cost an arm and leg back then. The analyst consensus back then was that OLEDs are too hard/expensive/etc for generic consumers.
- 5y ago
- toast0 5y agoI'm really hopeful that because these displays aren't (currently anyway) built as large planes and then cut to sizes, that we won 't be stuck with only the sizes and aspect ratios that cover the plane... I've got arcade machines that want 25" 4:3 monitors, and maybe MicroLED could fill the void. A suitable CRT would be best, but those are hard to find. Modern displays don't come in 4:3 at that size, and specialty 1:1 displays are too expensive for me to even measure to see if they'd fit.
- pedrocr 5y agoFor an arcade can't you get away with mounting the screen vertically by hiding the rest of it in the cabinet?
- toast0 5y agoOn some machines, yes. Not on a sit down machine where your legs go under the screen, and there's stuff on top of the screen. Maaaaaaaaybe, this machine could get by with an ultra-wide screen, because it's got two player stations next to each other with a partition in the middle, but that's a big maybe. Plus I'd need something to link the two independently clocked video feeds into a single wide display.
- hlandau 5y agoHow is light to be modulated with these? They're not going to use PWM, are they? I'm pretty tired of seeing displays that exhibit visible flicker when a camera is pointed at them, especially as flicker imperceptible to the human eye can still be migranogenic.
- ayane_m 5y agoThe effect largely goes away if the frequency exceeds 4 kHz, as even the quickest saccades would see a blur rather than a flicker. putting the flicker above 20 kHz would be ideal, because that way it is well beyond both visual range and even any noise generated by the circuit would be inaudible. The real question is why higher frequencies aren't used for driving LEDs. They have very low cycling latency, so it's a no-brainer.
- jiggawatts 5y ago> They have very low cycling latency, so it's a no-brainer. I read somewhere that that makes it difficult to accurately control the brightness. With HDR content already requiring 10 bits, anything that interferes with that is a problem. Future panels might require 12 bits or more.
- daniel_reetz 5y agoIn the mini-led industry the clocks are often much higher. That said, you are limited by scan (how many LEDs driven per driver chip), because the clock is divided among the driven LEDs. Consider that raising the PWM frequency doesn't automatically solve the problems you raise. 20khz PWM can still have audible harmonics, for example. In the mini-LED industry there are actually spread spectrum/randomized PWM approaches to address this, and to help with EMI.
- watersb 5y ago> 20khz PWM can still have audible harmonics... there are actually spread spectrum/randomized PWM approaches to address this, and to help with EMI. !! Great.. now I have to consider Tempest shielding for all the displays; I had thought once CRTs went away, we'd be safe...
- mensetmanusman 5y agouLEDs would let us read outdoors without brightness throttling that makes OLED displays invisible with sunglasses on. That would be a major improvement for mobile.
- TooSmugToFail 5y agoAny guesses who may be the first to market with a uLED micro displays?
- qilo 5y agoApple Watch?