240hz ULMB at 480hz monitor better than 120ULMB at 240 ?

Ask about motion blur reduction in gaming monitors. Includes ULMB (Ultra Low Motion Blur), NVIDIA LightBoost, ASUS ELMB, BenQ/Zowie DyAc, Turbo240, ToastyX Strobelight, etc.
Post Reply
rrisk93
Posts: 20
Joined: 05 Mar 2019, 01:22

240hz ULMB at 480hz monitor better than 120ULMB at 240 ?

Post by rrisk93 » 05 Mar 2019, 01:32

I understand 120hz ulmb at 240hz monitor is better than 120hz ulmb at 144hz monitor.

Then, 240hz ULMB at 480hz monitor will look better than 120hz ulmb at 240hz monitor?


And if you compare ulmb and dyac, which is better at removing blur from your experience?

User avatar
Chief Blur Buster
Site Admin
Posts: 11648
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: 240hz ULMB at 480hz monitor better than 120ULMB at 240 ?

Post by Chief Blur Buster » 05 Mar 2019, 04:34

There's no 480Hz monitor with 240Hz ULMB yet.
This is an unknown, there's no Yes/No answer to this. It's all in how the pixel response curve works.

But let's say, theoretically, let's do the math.

120Hz ULMB on 240Hz = 8.33ms refresh cycles with 4.16ms scanout, 4.16ms VBI
240Hz ULMB on 480Hz = 4.16ms refresh cycles with 2.08ms scanout, 2.08ms VBI

So the short answer is: Assuming pixel response is unchanged (1ms), then probably not.

120Hz ULMB on 240Hz = 8.33ms refresh cycles with 4.16ms scanout, 4.16ms VBI
240Hz ULMB on 480Hz = 4.16ms refresh cycles with 2.08ms scanout, 2.08ms VBI

You ideally need GtG(100%) to fit in the VBI, but usually that is not possible, so you have GtG(90%) occuring in VBI, which results in average strobe crosstalk found at http://www.blurbusters.com/crosstalk (e.g. a 10%-faint afterimage strobe crosstalk at the top/bottom). Even GtG100% takes more than a refresh cycle, so the whole screen will have slight crosstalk.

Some strobe tuning can help, see Animations of strobe tuning a BenQ monitor

The blanking interval is the time for GtG pixel settlement time + strobe flash.
You can see high speed videos of panel scan out at http://www.blurbusters.com/scanout
You can see high speed videos of ULMB-like mode at http://www.blurbusters.com/lightboost/video

So if GtG(100%) is much better on the 480Hz monitor, you could probably finish most of the GtG and flash the backlight in the very tiny 2ms blanking intervals. That's a tall order though.

As we start exceeding 480Hz, then there's really little reason to flash the backlight anymore because 480Hz itself is strobeless ULMB. ULMB is already roughly ~1.5ms persistence.

Basically ULMB reduces persistence of 120Hz from 8.3ms (1/120sec) down to approximately 1.5ms persistence already.
But if you can do 480fps at 480Hz, that's already 2ms persistence without strobing!

Image

Even 480Hz is not yet the final frontier, the mouse arrow is not yet a continuous blur.
And ULMB doesn't fix the stroboscopic effect.

Image

Doubling Hz and frame rate can halve motion blur. At some point, it can cease to become necessary to strobe when you've achieved strobeless ULMB or blurless sample-and-hold. Real life doesn't strobe. Strobeless ULMB is the Holy Grail, and that's why Blur Busters are huge mega-fans of 1000Hz monitors of the future (expected to arrive prior to 2030s).

It's a long-term slow Refresh Rate Race to Retina Refresh Rates. See Blur Busters Law And The Amazing Journey To Future 1000Hz Displays. No blur. No stroboscopic effect. No flicker. No strobing needed. ULMB achieved strobelessly. It's going to take years, but so did 4K. In less than two decades, 1000Hz displays will be affordable and practical -- if it's cheap enough to add to a display, and it benefits humankind, it'll happen eventually as people understand why we need ultra-Hz to fix a lot of problems.

ULMB = approximately 1.5ms persistence (though ULMB Pulse Width adjustment can make it as low as 0.25ms, albiet very dark)

Without strobing, full frame rate at a true native refresh rate of:
60Hz = 16.7ms persistence blurring
120Hz = 8.33ms persistence blurring
240Hz = 4.16ms persistence blurring
480Hz = 2.08ms persistence blurring
1000Hz = 1ms persistence blurring
2000Hz = 0.5ms persistence blurring

Image

Beyond about 240Hz, strobing becomes massively less important, and we instead use ultra-Hz as the method of blur reduction instead of strobing. Now you need high frame rate to eliminate motion blur strobelessly. For that, the GPU is a brick wall but that will be a solved problem; various frame rate amplification technologies can help increase frame rates, with work being done towards potentially huge ratios such as 10:1 ratios, e.g. converting 100fps to 1,000fps virtually laglessly -- hopefully within a decade.
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!

rrisk93
Posts: 20
Joined: 05 Mar 2019, 01:22

Re: 240hz ULMB at 480hz monitor better than 120ULMB at 240 ?

Post by rrisk93 » 05 Mar 2019, 05:45

wow Thanks!!

If we want ULMB 10% clarity with no brightness loss, we need 4000hz OMG

Maybe it can be faster to rise monitor brightness(HDR things) and apply ULMB than reach 4000hz Holy Grail...

User avatar
Chief Blur Buster
Site Admin
Posts: 11648
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: 240hz ULMB at 480hz monitor better than 120ULMB at 240 ?

Post by Chief Blur Buster » 05 Mar 2019, 18:16

Yes!

(Just one thing: ULMB 10% isn't the same thing as LightBoost 10%, since ULMB Pulse Width is a different scale and range than LightBoost % adjustment.)

There's the very old LightBoost 10% vs 50% vs 100% photo comparision page. too

Image

LightBoost 100% = approximately 2ms persistence
LightBoost 50% = approximately 1.5ms persistence
LightBoost 10% = approximately 1ms persistence

At a normal 960 pixels/second movement speed, better-than-1ms persistence will be hard to see. However, it's easy to see 0.5ms MPRT persistence versus 1.0ms MPRT persistence with faster motion speeds with tiny text: TestUFO Panning Map Test at 3000 pixels/second ...
You cannot read that with default ULMB! ... You must reduce ULMB Pulse Width lower until you see the street name labels.

As a rule of thumb, readability of 6-point text at "X" pixels/second requires "(2/X)" persistence. So 2000 pixels/second motion requires 2/2000sec persistence = 1ms persistence. The math is very simple. Just play with TestUFO Panning Map test, and the Blur Busters Law suddenly becomes obvious.

Blur Busters Law:

1ms of persistence translates to 1 pixel of motion blur per 1000 pixels/second


Blur Busters Law (which is simply a heavily simplified MPRT formula) is very easy to demonstrate with the ULMB Pulse Width setting with the TestUFO panning map test.

The diminishing curve starts really biting our ass as the curve becomes a cliff. Realistically, we can simply be happy with 1000Hz. That's going to be the sweet spot circa 2030 when monitor technologies and GPU technologies (frame rate amplification) converges into something affordable for mid-range.

I predict the first 1000Hz implementations will hit the consumer market approximately 2025.

That said, many people mis-guess the curve because of other things (e.g. fighter pilot tests or flicker tests) but they neglect the phantom array effect, the wagon wheel effect, and the persistence motion blur effect.

I've got a heavily upvoted answer on Quora about quad digit refresh rates still having side effects. The "Humans can't see beyond 300fps" is flat earth thinking when we don't consider all the variables that creates various side effects on a finite-refresh-rate display.

The jump from 60Hz to 120Hz is an 8.3ms blur improvement.
The jump from 120Hz to 1000Hz is a 7.3ms blur improvement.

Strobeless 60Hz = 16.7ms blur
Strobeless 120Hz = 8.3ms blur (8.3ms better)
Strobeless 1000Hz = 1ms blur (7.3ms better)

That's a giant jump up the curve-cliff to get a "wow, 120Hz->1000Hz is as big jump as 60Hz->120Hz" because the jump from "120Hz->240Hz" is so marginal.

The incremental approach to refresh rates means the Refresh Rate Race to Retina Refresh Rates is kind of a slow Moore's Law (refresh rates doubling approximately every 5 years since around year 2010). So because of the technology gradual improvement, it will be good for monitor manufacturer profits though since it's not easy to dramatically jump up the diminishing curve of returns.
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: 11648
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: 240hz ULMB at 480hz monitor better than 120ULMB at 240 ?

Post by Chief Blur Buster » 05 Mar 2019, 18:47

This is a good thread to also remind readers of the accomplishments of Blur Busters.

Usually we are modest, but it's time to toot/boast a little, because -- not everyone realize what we've done to the display industry behind the scenes. So I need to boast a bit, just to make sure readers don't dismiss the dream of strobeless lagless ULMB.

Accomplishments by Blur Busters in Nudging This Refresh Rate Race

We flip a lot of dominoes & we flap a lot of butterflies as a itty-bitty role in the Chaos Theory of the current refresh rate race -- indirectly effecting major change to the display industry, one itty nudge at a time. Most of these nudges nobody notices at first, but has had a big effect five years later.

We launched the Blur Busters Forums because NVIDIA gifted us a few GSYNC upgrade kits which we announced in this December 17th, 2013 forum post. Blur Busters was also the first website to discover how to test the input lag of GSYNC, in GSYNC Preview #2. In the process of all this, we found the VG248QE ULMB mode did not match LightBoost 10% quality, and we made sure NVIDIA solved this problem. So now all monitors with ULMB (after the VG248QE Upgrade Kit) now has the ULMB Pulse Width option -- thanks to Blur Busters! Blur Busters is the one that convinced NVIDIA to add a "ULMB Pulse Width" option.

Also, many display reviewers use our tests that had a peer reviewed conference paper with NIST.gov, NOKIA, and Keltek. Any time you see our UFO in a display reviewer (like RTINGS or TFTCentral or TomsHardware or PCGamer) or by a YouTube star, that's because I've given them permission to use the Blur Busters tests. Monitor manufacturers notice that, and they now use TestUFO as part of their tests. (TestUFO has literally been a free gift to help display industry innovation. Even the Oculus Rift VR kickstarter were one of the beta testers of TestUFO six months before it launched too!)

And many heads turned when I was the first person to publish tests of a 480Hz monitor. And the revelation of my formerly undisclosed involvement in convincing the VR industry to go low persistence. And now, I have contracts regularly with display manufacturers to help tweak their displays before launch (things like overdrive tuning, strobe tuning, etc). So rest assured, Blur Busters actually effect change to the monitor industry. We may be a media site at the cover page, but we now have a laboratory and consulting service too.

1000Hz might be expensive like 4K was twenty years ago, but remember today 4K can be purchased for cheap in Walmart. Ultra-Hz doesn't have to stay expensive for a century. Obviously, the engineering has to pay for itself so it'll be a slow Refresh Rate Race. Remember, 120fps at HD used to be a pipe dream.

We can do 1000fps from a GPU cheaply by 2030 or 2040 using various frame rate amplification technologies. People have stopped hand-wave-offs and laughing thanks to Blur Busters. VR scientists, AMD scientists, NVIDIA scientists, employees as some monitor manufacturers, etc, are all often big fans of the Blur Busters advocacy now (often secretly).

The usefulness of ultra-Hz becomes immediately useful with the education of staring at TestUFO tests on any high-Hz display in "see for yourself" (even the grandma agrees they can see the difference). Now, imagine if ultra-Hz becomes affordable (GPU wise and display wise). Someday. Within this human generation.

The starting pistol of the Refresh Rate Race began with the 2010s which Blur Busters launched early in. Throughout 2020 and beyond, advocacy of 1000Hz will ramp up. We may not be bigtime players, and we're only little components, but we tip a lot of dominoes and flap a lot of butterflies in the Chaos Theory of this Refresh Rate Race.

Blur Busters definitely exists to help accelerate the Refresh Rate Race to future Retina Refresh Rates at the both ends of the candle (the display side and the GPU side). We make sure we light the fire under GPU manufacturers and monitor manufacturers. ;)
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