[Overcoming LCD limitations] Big rant about LCD's & 120Hz BS

Advanced display talk, display hackers, advanced game programmers, scientists, display researchers, display manufacturers, vision researchers & Advanced Display Articles on Blur Busters. The masters on Blur Busters.
User avatar
Chief Blur Buster
Site Admin
Posts: 11647
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: [Overcoming LCD limitations] Big rant about LCD's & 120H

Post by Chief Blur Buster » 12 Aug 2017, 13:47

thatoneguy wrote:I still stick by my theory that if we can achieve a lower image persistence than the human eye can perceive(say for example 1 nanosecond) the multiple image effect will be eliminated at any refresh/framerate combo.
It's actually pretty close, but not identical.

Persistence affects eye-tracking.
Multi image affects non-eye-tracking (stationary gaze).

-- Imagine a theoretical robot arm moving a theoretical 1,000,000 Hz mouse.
-- Imagine the robot arm moves the mouse really, really, really fast -- like 1,000,000 pixels per second.
-- You do a fixed gaze.
-- On a theoretical 100,000 Hz (100KHz) refresh rate display -- the mouse arrow steps would be 10 pixels apart (1,000,000 / 100,000 = 10)
-- That said, with the refresh cycles only 1/100,000sec you will need a huge number of nits (e.g. 10,000 nits or 100,000 nits) for white pixels) in order for the 100KHz phantom array effect to be really visible
-- Indeed, such a display would be too blindingly bright (even if only the mouse arrow pointer was that insanely bright).

(Note: An infinitesimally short flash is visible to the human eye, given sufficient photons. For example, a 1 microsecond flash is visible to human eye if it's 1000 times brighter than an equivalent 1 millisecond flash. Same number of photons. Equally "instantaneous looking" to human eye. But both are still visible.)

The solution to make both equal is:
(A) Enough hertz to solve phantom array effect for movements matching maximum human eye-tracking speed
This fixes eye tracking situation
(B) Add GPU motion blur between refresh cycles to eliminate the phantom array effect
This fixes stationary gaze situation

Doing (B) will slightly interact with (A), so you might need to go to a 2x factor above maximum to simultaneously satisfy (A) and (B). Basically X,XXX Hz with GPU-added motion blur of 1/X,XXXth second between all adjacent refresh cycles.
Head of Blur Busters - BlurBusters.com | TestUFO.com | Follow @BlurBusters on Twitter

Image
Forum Rules wrote:  1. Rule #1: Be Nice. This is published forum rule #1. Even To Newbies & People You Disagree With!
  2. Please report rule violations If you see a post that violates forum rules, then report the post.
  3. ALWAYS respect indie testers here. See how indies are bootstrapping Blur Busters research!

User avatar
Chief Blur Buster
Site Admin
Posts: 11647
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: [Overcoming LCD limitations] Big rant about LCD's & 120H

Post by Chief Blur Buster » 12 Aug 2017, 13:57

Also good new TestUFO animation, relevant to subject: http://www.testufo.com/displaymotionblur

1. Look at the stationary UFO. Observe vertical lines.
2. Look at the moving UFO. Observe that an image appears (easier to see on sample-and-hold displays; LCDs or OLEDs, not CRT)

Please turn off Blur Reduction / strobing / ULMB / LightBoost for this animation
For best results, use a desktop display, rather than a mobile display



Doubling your refresh rate will double the clarity of this image (twice the horizontal resolution).

The image never becomes fully sharp/clear until you reach a Hz equalling pixels per second. For 960 pixels per second, you need a true 960 Hz display (CRT, LCD, OLED) for http://www.testufo.com/displaymotionblur to match the clarity of http://www.testufo.com/photo

Blur Reduction (CRT impulsing, LCD strobing, OLED rolling scan) is ultimately only a very good band-aid in the humankind invention of static images to represent moving mages. Long term (in a decade, in a century, wide-FOV monitors, VR, Holodeck, etc) we need framerateless displays or ultra-high-Hz displays.

This new TestUFO animation [First time ever done on a computer screen, AFAIK] is also based on well-known optical principles that spinning LED bike wheels and old-fashioned 1920s/1930s mechanical televisions (Nipikow wheels), which used a single flashing neon or flashing LED running at thousands of Hertz to generate an image. More information I've posted about this funky new TestUFO test. What this means, is flashing a pixel faster, can actually increase resolution of effects like these. This is relevant to occlusion effects in real life.

Real world applications of ever-higher-Hz: Rapid occulsion effects. Picket fences. Cracks through doors. Etc. Playing FPS games while looking through dense bush or cracks or fence slats. Or trying to peek through a very tiny crack in a door in a FPS game (millimeter-league cracks that are only a single or two pixels on screen) via turn/strafing left-right to "rapidly scan through the crack". (Admit it, you might even scan-peek through the crack of a bathroom stall door sometimes to check if a toilet stall is occupied!). In real FPS games the cracks are much bigger but let's assume a much tinier crack that are only a few pixels wide, and you're wanting to "scan" faster. The rapid occulsion-deocculsion effects look better the higher the refresh rate you go. This is increasingly important the closer in humankind, that we try to approach Holodeck realism during virtual reality situations. Fixing occulsion-effects limitations fully will require multi-thousand-Hertz to make things look analog-motion even in these types of occlusion scenarios (which are not interpolatable), and pass the "Holodeck Turing Test" in the "Wow, I didn't know I was wearing a VR headset instead of wearing a transparent ski goggles" type of blind-testing that would definitely require ultrahigh-Hz retina-resolution headsets... While it's nitpicking right now, it still underscores the need for ultra-high-Hz to approach Holodeck-quality analog motion with no visible side effects...

I invented this TestUFO pattern in part because of this thread. More TestUFO demonstrations are on their way to scientifically demonstrate how higher Hertz still improve various specific kinds of image effects.
Head of Blur Busters - BlurBusters.com | TestUFO.com | Follow @BlurBusters on Twitter

Image
Forum Rules wrote:  1. Rule #1: Be Nice. This is published forum rule #1. Even To Newbies & People You Disagree With!
  2. Please report rule violations If you see a post that violates forum rules, then report the post.
  3. ALWAYS respect indie testers here. See how indies are bootstrapping Blur Busters research!

thatoneguy
Posts: 181
Joined: 06 Aug 2015, 17:16

Re: [Overcoming LCD limitations] Big rant about LCD's & 120H

Post by thatoneguy » 18 Aug 2017, 23:44

Chief Blur Buster wrote:Long term (in a decade, in a century, wide-FOV monitors, VR, Holodeck, etc) we need framerateless displays or ultra-high-Hz displays.
Personally I think the key to achieving this is getting to low enough persistence that humans can't detect. I think the end goal for display technology(whether that'd be TV's, Monitors, VR etc.) is completely fooling the human eyes and making it impossible for them to distinguish display footage from reality. For that we need lower persistence than the human eye can detect.

I think it will probably come in the form of extremely high refresh rate displays(10000hz or 100000hz to combat brightness loss) with extremely low persistence(nanosecond or picosecond-tier persistence) and extremely high brightness(possibly 100000 nits).

As for framerateless displays I still can't wrap my head around them :lol: :lol: :lol:

User avatar
Chief Blur Buster
Site Admin
Posts: 11647
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: [Overcoming LCD limitations] Big rant about LCD's & 120H

Post by Chief Blur Buster » 19 Aug 2017, 22:20

thatoneguy wrote:I think it will probably come in the form of extremely high refresh rate displays(10000hz or 100000hz to combat brightness loss) with extremely low persistence(nanosecond or picosecond-tier persistence) and extremely high brightness(possibly 100000 nits).
You only need extra brightness to compensate for black periods.

If you eliminate black periods by using ultra-high-Hz sample-and-hold instead, then brightness is NOT a problem.

You do not need extra brightness for low persistence achieved via ultra-high-Hz(& fps) instead of strobing

If you successfully avoid strobing/impulsing, 1000 nits is the same regardless of 60Hz, 120Hz or 10,000Hz.

Extra brightness is only needed to compensate for black periods. If you're doing 90%:10% BFI where it's dark 90% of time, and bright 10% of the time, you need 10x as much brightness in that 10% of the time. But it's still essentially the same total number of photons per second -- just crammed into 1/10th of the time because of the black period between strobes. The shorter the flash, and longer the black periods, the brighter the flash must be to compensate for the black periods, to net in the same average brightness.

1/10th BFI ratio (90%:10% OFF:ON) = 10x brighter flash needed for same average brightness as non-strobed
1/100th BFI ratio (99%:1% OFF:ON) = 100x brighter flash needed for same average brightness as non-strobed

But if you achieve low persistence via ultra-high-Hz, you don't have strobing, then you don't need brighter at all.

0.1ms persistence can be achieved by 0.1ms strobe backlight flashes. But you can also instead achieve 0.1ms persistence via 10,000fps@10,000Hz too (essentially blurfree sample-and-hold!). The diminishing points of returns finally end (for theoretical max-retina + max-FOV; a Holodeck) somewhere in the quintuple-digit Hz. But for a desktop monitor, it's somewhere in the quadruple-digit Hz. LightBoost 10% is 1ms persistence and LightBoost 100% is 2ms persistence -- see LightBoost 10% versus 50% versus 100% -- several of us still see the difference. Mathematically, to achieve 1ms persistence without strobing, you need all frame visibilities to be 1ms each. Which means 1000fps@1000Hz.

However, alternatives might happen. For camera recording, a workaround to avoiding discrete refresh rates (and its side effects), is photon cameras. Essentially timecoded photon recording / playback or other solution. They exist for other applications (e.g. this) -- it's essentially a framerateless camera that exists! Simply precisely timecoding the arrival of each photon at each pixel. Then you'd simply have a display that plays back the photons as accurately as possible -- the higher the Hz a display, the more accurate the playback is (fewer and fewer side effects such as unwanted motion blur). Unless humankind invents a framerateless analog-motion display, of course.

On topic, but quoted from the 480 Hz monitor test results:
Chief Blur Buster wrote: The Blur Busters Law:
"1ms of persistence (frame visibility time) equals 1 pixel of motion blur during 1000 pixels/second motion."

For strobed displays, persistence is the strobe flash length.
Changing refresh rate does not change the motion blur seen on strobed displays.

For CRT displays, persistence is the effective visible phosphor decay time.
Changing refresh rate does not change the motion blur seen on a CRT.

For sample-and-hold (LCD, OLED) displays, persistence is the refresh cycle length.
Doubling the refresh rate can halve motion blur seen on an ideal sample-and-hold display.
Head of Blur Busters - BlurBusters.com | TestUFO.com | Follow @BlurBusters on Twitter

Image
Forum Rules wrote:  1. Rule #1: Be Nice. This is published forum rule #1. Even To Newbies & People You Disagree With!
  2. Please report rule violations If you see a post that violates forum rules, then report the post.
  3. ALWAYS respect indie testers here. See how indies are bootstrapping Blur Busters research!

Post Reply