Chief Blur Buster wrote:...Blacks are very good for horror games. For many, it may be preferable over other OLED artifacts.
However, the new LCD VR has higher resolution and less motion blur, so many others prefer that too. Quest 2 has virtually no screendoor effect, and its deliciously perfect zero strobe crosstalk (some incredible LCD engineering there!). And there's no human-detectable motion blur even during fast head turns, thanks to the current LCD VR having about 1/6th the motion blur of OLED VR at the moment.
As you already know, Blur Busters works with display manufacturers from time to time, so I can give you some inside information on why OLED has become problematic for VR.
This is because the LCD backlight flashes six times quicker (0.3ms flash backlight) versus the OLED pixels flashing for 2ms flash. The problem with direct-flashing of OLED pixels is large degradation in black noises/speckle the briefer you try to direct-flash an OLED pixel for. The shorter you try to flash an OLED pixel for, the worse the noise/speckle in darks become -- so 2ms was a compromise (and speckle is still visible).
In low-persistence engineering, LED / OLED speckle is a major problem for the bottom end of dynamic range during low persistence operation of an OLED -- and VR is aggressively pushing motion clarity requirements. There is a very dastardly vicious cycle effect (higher resolution & wider FOV & higher refresh rates = amplifies visibility of motion blur = 0.5ms MPRT versus 1.0ms MPRT is now human-visible when we're entering the leagues of strobed 4K VR displays). The way higher resolutions amplify the needs of motion clarity, is a huge driver of needing sub-1ms-MPRT for virtual reality headsets, and only the new LCD VR headsets (Index, Rift S, Quest 2) are doing a stellar job honest real-world measured sub-1ms-MPRT (non-fake).
Unfortunately, a law-of-physics effect called Talbot-Plateau Law (need to strobe brighter briefer) is causing major problems with low-persistence OLED.
Trying to brighten the OLED and flash it briefer -- also unfortunately brightens the speckle effect (the noisy-OLED effect during dark situations) every time we try to improve MPRT with an OLED (without using low-persistence sample-and-hold). Also, OLED GtG (0.1ms) starts becoming a significant percentage of a OLED low persistence (2ms). Whereas, the backlight flash is completely decoupled from LCD GtG (which can simply be hidden in the total darkness between refresh cycles). As long as you get GtG100% perfect between refresh cycles, LCD has no bottom limit to motion clarity -- it can be infinitely clear (e.g. 0.001 microsecond flash, if such a backlight existed). Surprisingly, LCD VR LCD motion clarity now currently beats CRT motion clarity, and Quest 2 is now being used with Virtual Desktop to emulate a 60 Hz CRT; running MAME or RetroArch on a floating virtual monitor (or virtual monitor sitting on virtual desk) while wearing Quest 2.
TODAY:
As already shipping in Index / Quest 2 / Rift S -- the Talbot-Plateau law is less of a barrier with LCD (outsourced light) than OLED (tiny pixels), since it's easier to design a heatsinked/watercooled LCD backlight that can flash stadium-bright like a CRT electron beam, to compensate for the briefness of low-persistence. This is hard to do directly with OLED pixels without degradations and extra speckle/noise in blacks.
FUTURE:
The good news is I saw a FALD VR prototype so LCD VR will have perfect blacks come ~2025. As long as it's backlit by at least 10,000 LEDs to 50,000 LEDs (or even 100,000), the bloom will be a nonissue. Historically FALD was scarily expensive, but prices will fall fast on MicroLED / MiniLED panels. It will probably follow a commoditization path much like 32x32 Jumbotron LED panels are under $10 off Alibaba -- that's 1024 RGB LEDs matrixes for under a penny per LED -- places like Alibaba
https://www.alibaba.com/showroom/32x32-led-matrix.htm
So FALD (Full Array Local Dimming) isn't necessarily expensive forever, even for VR headests. Probably Valve Index 2 will use them, come 2022-2023, maybe even as early as 2021. Hopefully CES 2021 will announcd FALD VR headsets, but that might not happen till CES 2022 due to COVID. It might be less than 10,000 LED FALD, but it's a start once FALD VR comes out.
It might not be until a few iterations later where FALD VR is cheap enough to arrive in a future Quest (maybe Quest 4 in 2025?) since Quest is hitting the low-cost VR market (and doing a stellar job at it). FALD LCD VR will be initially a premium feature. So, I think a theoretical future Valve Index 2 would probably do FALD VR first. Also, it would be neat if Index 2 has a "Lighthouses-optional" mode of operation. In other words, having both inside-out tracking and sensor-tracking modes of operation. Easy setup by making Lighthouses optional, with optional accuracy-improvements (for controllers-behind-back operations, etc).
Longer term, direct-view MicroLEDs will be superior to both OLED and LCD, but trying to have 33 million LEDs in an 8K VR headset headset is a bit unobtainium. It's MUCH easier to have a 1-million-LED monochrome MicroLED backlight driving ultra-fast-switch LCD that's successfully strobe-tuned to GtG100% perfection for the entire 256x256 GtG heatmap or 1024x1024 GtG heatmap.
Now, if "
low-persistence sample-and-hold" (1000fps+ on 1000Hz+) arrives to OLED, it solves the speckle problem overnight. Ultra high framerate OLED would do a really stunning job of blacks (much more speckle-free blacks than today). However, it will be quite a while before 1000fps 1000Hz becomes a reality, so for now we'll need low-persistence via strobing for now.
(P.S. Those lovely GtG heatmaps you see on HardwareUnboxed or ApertureGrille or RTINGS? Those are actually puny to what I bet John Carmack did -- i.e. ordering a massive GtG heatmapping & tuning the LCD VR panels did -- figuratively, imagine a spreadsheet of over 65000 GtG values with every single GtG value tuned. Every single GtG value of them fixed in Quest 2 at faster than a refresh cycle, with strobe flash perfectly timed as the pixels pass the exact GtG value desired. John Carmack magic. It's an engineer standing ovation. Imagine a 65,536 square HardwareUnboxed GtG heatmap with every square GtG100% virtually perfect at strobe time). The main flaw of that Quest 2 LCD is it's not backlit by a >100,000-LED-element-count FALD MicroLED panel, bloomfree and perfect blacks.)
_____
Knowledge is power; understanding how display motion blur helps understand the state of OLED VR technology and LCD VR technology. So even the peripheral topic helps educate, LCD VR versus OLED VR, and how different strobe flash lengths affects display motion blur -- is quite useful in understanding the pros/cons. Necessarily, talking about display motion blur, creates necessary topic sidetracks (
why strobing/flicker is currently necessary for VR, because of the impossibility of the refresh rates necessary to fix motion blur strobelessly). I do go above-and-beyond to educate people about how display motion blur physics works -- but my topic sidetracks are (usually) generally well-loved by advanced users. Some people cannot use VR because the person is too flicker-sensitive. Also, display motion blur is extra blur forced on you above-and-beyond your natural human limitations. Holodeck should perfectly match real life.
Bottom line, dsplay motion blur is unwanted extra motion blur above-and-beyond real life, which VR is attempting to simulate.
So, today, if your #1 priority is OLED blacks at high resolution -- buy a very good OLED headset such as HTC Vive Pro.
If your #1 priority is motion clarity, then modern LCD VR is superior -- Valve Index, Rift S, Quest 2
If you want excellent ease-of-use thrown in too, then the answer is a no-brainer (if you can tolerate Facebook) -- Quest 2