480 Hz Monitor Prototype Now at Blur Busters!

Posted Aug 10, 2017 by in: Exclusives

EXCLUSIVE! We have received a prototype 480 Hz monitor at Blur Busters. Sent to us in kit form, we are now currently testing — and an article is coming out shortly.

Follow Blur Busters on Twitter and Facebook, to watch for our upcoming article!

We are the world’s first website to show pictures of the world’s first full-color 480 Hz capable LCD monitor that is not a laboratory display (e.g. Viewpixx 1440Hz DLP).

If you simply cannot wait, check out our recent extensive 240 Hz input lag tests in the meantime!

To help support the relaunched Blur Busters, please SHARE this page.

Thanks,
Chief Blur Buster

About Mark Rejhon

Also known as Chief Blur Buster. Founder of Blur Busters. Inventor of TestUFO. Read more about him on the About Mark page.

27 Comments For “480 Hz Monitor Prototype Now at Blur Busters!”

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Sort by:   newest | oldest | most liked
Mark Rejhon
Author
Mark Rejhon

Tests still under way.

The great thing is that it does look like strobeless LightBoost. I’ve never seen that on any LCD before. And — this is just an off the shelf 28″ 4K LCD 1ms TN panel. With a custom motherboard+TCON. It isn’t a new panel at all. [That’s the jawdrop]. Probably, most manufacturers don’t even believe it’s necessary to get that far in refresh rate — and the low-resolution limitation is probably an issue for many.

But it makes “strobeless LightBoost” motion clarity possible. 480fps@480Hz looks like LightBoost without any strobe flashing. [Except for ghosting artifacts — there is no overdrive feature available]

Motion blur equivalence:

Persistence of strobed display = flash length (2ms on LightBoost)
Persistence of sample-and-hold display = refresh cycle length (2ms on 480fps@480Hz).

Strobeless motion clarity — CRT clarity in a sample-and-hold display will require going to 1000fps@1000Hz in the future (to equal 1ms MPRT strobed displays), and I now think 1ms TN LCD may actually eventually be capable of that with some ultra-high-quality overdrive (the type of overdrive calibration that is done by NVIDIA’s wizards. They do a stunning job of their ultra-precise overdrive calibration in some ULMB/LightBoost monitors, with photodiode oscilloscopes measuring >80% accuracy in some “1ms GtG” implementations — though the worst 240Hz displays do have twice as much motion blur than the best 240hz display — all because of overdrive lookup-table calibration differences.

I wish OLED could come sooner and do that sooner — but hey, if we’re stuck with TN — then at least, it looks like a well-tuned overdrive (80% accuracy in 1ms GtG) can provide good reduced-dynamic-range 1000fps@1000Hz (800:1 contrast instead of 1000:1) with little to no visible ghosting (like the world’s best 120Hz/240Hz monitors with the absolute-best calibrated overdrive).

I would surmise, that monitor makers will eventually be able to achieve 1000Hz gaming monitors within 10 years to achieve “strobeless ULMB” / “strobeless LightBoost” equivalence. As a result of this 480Hz display [even without overdrive yet], and photodiode oscilloscope measurements tested on some of my better 120/144/240Hz displays, I’m now convinced there’s still a lot of room to milk left in TN technology — and even beyond 480Hz (eventually).

Like3Vote Down 
6 years 8 months ago
POVHFR Videos
Member
POVHFR Videos

Wow! Simply amazing…
I would say it could be sooner than 10 years. My guess is that if there is more and more innovation and breakthroughs, based on what you said we could see 4k or 8k 1000hz – 4000hz if not more in 10 years (2027)

Like0Vote Down 
6 years 8 months ago
Mark Rejhon
Author
Mark Rejhon

Baby steps.

Consumer true-1000Hz by sometime during the 2020s (read: 2020-2029) is now starting to look feasible to me as a prediction. But it might be low-resolution and/or poor-gamut (e.g. TN LCD or some other tech that makes sacrifices).

It’s also possible, by then, OLED might be a savior, with some of the rolling-scan ideas I’ve written about in the Custom OLED rolling scans thread.

Like1Vote Down 
6 years 8 months ago
POVHFR Videos
Member
POVHFR Videos

I agree, baby steps. And yes, consumer true-1000hz should be feasible I would hope by 2021.
I agree that there would be another display by 2020-2029 that would make sacrifices but that is still in a way better than current 2017 technology. I can see how OLED during the next decade would be a savior.

Like0Vote Down 
6 years 8 months ago
POVHFR Videos
Member
POVHFR Videos

At 480hz, how clear is the motion compared to reality. Is the fluidity smooth enough to trick the eye into thinking that it’s an actual moving object?

Like1Vote Down 
6 years 8 months ago
Mark Rejhon
Author
Mark Rejhon

Not yet. Reality will require at least about 10,000 Hz (assuming retina rez + 180 degree FOV);

A good demonstration motion test of this principle is http://www.testufo.com/motionblurdemo

Also, look at the mouse-arrow image above. If you move your mouse at 960 pixels per second, you need 960Hz to fill every spot with a mouse arrow.

Mouse effect during 60 Hz vs 120 Hz vs 240 Hz

There are many side effects caused by the artificial human invention of using static images to represent moving objects. During fixed gaze, you got stroboscopic stepping effect. During eye tracking, you got motion blur issues. Neither low Hz nor impulsing (CRT, LightBoost, OLED etc) nor GPU blur effects (fix stepping effect) can simultaneously solve both at the same time.

Even Michael Abrash, the chief scientist of Oculus, have said this:

“…the sweet spot for 1080p at 90 degrees FOV is probably somewhere between 300 and 1000 Hz, although higher frame rates would be required to hit the sweet spot at higher resolutions….”

Modern display researchers now agree. NVIDIA has tested a 16,000 Hz augumented reality prototype, and Viewpixx has a 1440Hz laboratory DLP projector.

The bigger the FOV and/or the more pixels, the more visible the issues are. Bigger displays, retina desktop displays, VR headsets too. More rez + more FOV = more pixels to track eyes over = more Hz needed for matching reality. The Hz limiting factor includes eye tracking speed.

Diminishing points of returns? Yes.
Vanished? No, not till quadruple-digit / quintiple-digits Hertz.

Like2Vote Down 
6 years 8 months ago
POVHFR Videos
Member
POVHFR Videos

Wow! I mean, I’m a videographer, I record videos at 720p 240fps, and wanted to know how fast to record and how fast of hertz I would need the monitor to refresh to be as clear as the real world. I thought that 240fps would be enough… You are saying I need upwards of a “modern day” slow motion camera, where the camera could take video at 960fps to 2000fps? The monitor in the near future perhaps would need to refresh at 1000hz (1khz) to 16,000hz (16khz)? Which hertz rate would be more realistic in the next two or three years (2020-ish) for completely clear visuals?

Like0Vote Down 
6 years 8 months ago
POVHFR Videos
Member
POVHFR Videos

And I can’t wait to see the emerging 480hz displays, should be very exciting!

Like0Vote Down 
6 years 8 months ago
Mark Rejhon
Author
Mark Rejhon

Very neat that you’re doing 240fps video! Are you playing your 720p 240fps videos in real-time on a 240Hz monitor?

You don’t really need to go much higher than high a frame rate if your resolution is only 720p.

Resolution is a limiting factor in clarity benefits of Hz, so you need higher resolution for the higher Hz to be really dramatically worth it. 1280×720 at 240Hz means one screen width panning in one second only has 5 pixels of motion blurring (720/240 = 3), assuming camera shutter didn’t add extra panning motion blur. It’ll still look pretty damn good, as video motion blur isn’t really objectionable (We don’t hate motion blur all the time — it has artistic purposes too — we just prefer we have the ability to *turn off* the motion blur for many situations.) Often high speed video will look less clear at the same resolution, so you likely got more sensor blurring from clarity limitations enforced by high speed video.

Now, if you’re filming at 3840×2160 at 240fps, you will have far more sharpness difference between static images and panning images. 3840/240 = 16 pixels of display motion blurring at 3840-width for 1-screen-width-per-second panning. Now if you have a 1/500sec shutter, that also adds an additional 3840/500 = ~8 pixels of motion blurring during 1-screen-width-per-second panning (3840 pixels/second). Now your cumilative motion blur is approximately (16+8) = 24 pixels of law-of-physics absolutely guaranteed minimum panning motion blur on a 3840×2160 sample-and-hold display with instant 0ms GtG.

People not familiar with this, can obtain a good model of a 240Hz eSports gaming monitor, and staring at the 30fps-60fps-120fps-240fps motion comparision, to realize the beautiful mathematical relationship of “1ms of frame visibility time = 1 pixel of motion blurring for every 1000 pixels/second”. (Excluding additive blurring/ghosting from GtG limitations). Double frame rate very clearly & reliably halves the motion blur of panning scenery on a sample-and-hold display. (For video; assuming no shutter-based motion blurring and no GtG-based limitations added). But as you can obviously see from the motion blur math formula, 240Hz isn’t enough to look perfectly like reality if we’re dealing with ultra-high-resolution videos (e.g. 4K or 8K sensor with no bayer-induced blurriness and no highspeed-camera-induced blurriness).

You’ll only need quintuple digit Hz if you’re filming for a 180-degree retina-resolution strobeless display (e.g. 16K video — or maybe even 32K video to add a 2x oversampling factor to compensate for Bayer sensor blurriness, etc). Ouchie.

Unfortunately, yes it is a vicious circle: The higher the resolution your video is, the higher the frame rate you will need to eliminate the “panning blurring effect” on sample-and-hold displays (like an LCD matching the frame rate of the video).

Think of it this way, 1ms of frametime = 1 pixel of motion blur during 1000 pixels/second motion

Which means 240fps@240Hz has half the motion blur of 120fps@120Hz (Assuming LCD GtG isn’t the limiting factor). Jumping from say, 120Hz to 144Hz to 165Hz is so small, but you need to jump progressively bigger as you go up — e.g. 120Hz -> 240Hz -> 480Hz -> 960Hz just to halve motion blur every single time.

So if you have panning video scenery at 4000 pixels/second (4K video, 1 second from left edge to right edge), you’d want 4000fps@4000Hz to keep motion blur to only 1 pixel thickness (aka CRT clarity, but without strobing or flicker).

So, for now, we’re stuck with impulsing (CRT impulsing, LCD strobing, OLED flicker rolling scan) as a band-aid solution to keep video panning scenery as crystal-clear as stationary scenery. You can easily do that at just a mere 60fps combined with a fast per-frame shutter (e.g. 1/1000sec shutter).

So your weak links for motion blurring include:
– Display-end: Frame visibility time (persistence, sample-and-hold), pixel response limits (transitions, GtG), resolution (low-rez thru retina)
– Camera-end: Shutter speed time, frame resolution
– Human-end: Eye tracking speed & eye resolution.

They interact. For example, the faster video imagery pans, the more motion blur there will be during video. If you have a 1000Hz display (1ms frame visibility time), a 1/1000sec shutter (1ms video motion blur time). it still additively adds up to approximately 2 pixels of motion blurring because you’re adding together the camera-end weak link and the display-end weak link. Double ouch. Now you can try and fix the display-end motion blur limitation by simply flickering the display (like a CRT, or a Blur Reduction LCD, or a scanning backlight LCD), or adding frame interpolation (fake frames, Motionflow, etc). That works to an extent but has side effects too, and doesn’t fix blur from slower shutter.

I suspect that to record life in a true “Star Trek Holodeck” manner, we’ll probably need cameras that can record geometry (whole scenery holographically in 3D, essentially) at a ultra-high-resolution. And stored in an analog framerateless manner. Basically, a method of storing infinite framerate via vector mathematics, etc. And having it automatically adapt to the refresh rate of any display. This is a “far future thinking” (10years, 100years?).

To pass a theoretical Holodeck Turing Test (“Wow, I did not know I was wearing VR headset rather than transparent ski goggles!”), you’re looking at human eye tracking speed as the limiting factor, as well as solving this kind of effect (see New Motion Blur Demo that even a 240Hz display can’t successfully fully resolve). If you already have a 240Hz display, change your refresh rate between 60Hz, 120Hz, and 240Hz. Now imagine a 4K display, and run at a motionspeed of 3840 pixels/second, and you now begin to realize the necessity of 4000fps@4000Hz for “it looks like real life” on a 40-degree FOV 4K display (approx). Now, if you’re going Retina 180-degree FOV, (theoretical Holodeck), the refresh rate minimums to pass the reality turing test, jumps up into the quintuple digits.

But is it important? Not really. Displays don’t need to match reality perfectly. But Holodecks will need to.

However, as a rule of thumb: The higher resolution + the wider the FOV = the more Hz you need

Real life doesn’t flicker or strobe like a CRT. So you need continuous illumination. And that’s…. The dreaded sample-and-hold. And that means, lots of motion blur even at instant pixel response (0ms GtG)

– If your shutter speed is slow, Hz becomes less important as camera shutter speed is limiting factor
– If your resolution is low, Hz becomes less important as clarity of images won’t change as much (stationary vs panning)
– If your panning speeds are low, Hz becomes less important too.
– However, if you’re wanting to hit the limits….

…Then you’re targetting the limits of human eye tracking speed (more Hz needed) to try to make panning video exactly as sharp as stationary video (+even more Hz needed), on a Retina display (+even more Hz needed) that’s very wide-FOV such as 180-degree surround vision (+more Hz needed), you stack on a lot of multiple simultaneous pre-requisites that stack up to needing quadruple-digit Hz displays (and beyond in the case of 180-degree retina Holodecks — that requires quintuple digits to match reality. The limiting factor is human eye’s tracking speed while still being able to pay attention. Seeing something suddenly defocus when things pan, is very noticeable. Whether you’re doing panning video scenery, or if you’re doing head-turning in virtual reality. The sharper the original stationary image, the easier to detect the defocussing effect of panning motion blur (including motion blur from Hz limitation on a sample-and-hold display). The wider the FOV, the more time your eyes has to track.

Yes, it’s a vicious circle — the higher the resolution & the bigger the display — the more Hz you need to make it perfectly match reality (aka looking-out-of-a-window effect). This pretty much guarantees an extremely long technological progress ahead of us, even if details doesn’t really matter to most users today who are perfectly happy with 120Hz or 240Hz.

Display engineers can see the Area 51 at https://forums.blurbusters.com to understand the science behind various motion blur principles — and other side effects of the humankind invention of flipbooking a series of static images to represent analog moving images. We have a very, very long technological progress path ahead of us before the $50 Apple-Oakely “Virtual Reality Sunglasses” from Apple Store in year 2050 — will look as good as reality (ala Star Trek Holodeck). Who knows what mainstream gaming will be in about 50 years from now? We cannot let Hertz progress stop, because people falsely think “1000Hz doesn’t matter”.

Needless to say: I’m impressed that you’re even working on 240fps HFR realtime-playback video! (if I hear you correct).

Like1Vote Down 
6 years 8 months ago
POVHFR Videos
Member
POVHFR Videos

Thanks again!! You heard me correct, 240fps realtime playback video :))

And I’ve been working on High Frame Rate videos at 60+fps since 2011, and triple digit fps since 2015. I have a laptop that I overclocked to 108hz that does look pretty good, but obviously there are still visible scattered frames when viewing my content on it at 108fps.

Like0Vote Down 
6 years 8 months ago
Mark Rejhon
Author
Mark Rejhon

Doubtful we’ll be able to progress sufficiently far to get simultaneously ludriciously high refresh rates AND high resolution by the 2030s. But who knows?

Progress is fast, but Moore’s Law lost a lot of steam. Scaling resolution and refresh rates simultaneously is hard. Some math calculations for retina VR looking like reality, would end up having a display requiring tens of trillions bits per second (16K @ 30bit @ 10KHz). Solutions such as rendering a higher-resolution at eye gaze center could be very helpful in this long technological progress, as peripheral vision does not need to be insane resolution. This works for VR use case, but doesn’t work for video because someone might stare at a differents parts of video, so every location in video would have to be retina-clear to look like reality.

So that’s where future 3D games may have an advantage: They can “cheat” by only rendering the most detail at eye gaze centre instead of peripheral vision (e.g. future eye-tracked VR displays with eye-tracking-compensated 3D rendering)

Baby steps first though. I now see 1080p 1000Hz feasibly happening by the 2020s (read: 2020-2029).

For now we’ve got 240Hz displays and emerging 480Hz displays to contend with.

Like0Vote Down 
6 years 8 months ago
POVHFR Videos
Member
POVHFR Videos

Okay. I re-read all of your comments over again, and so if I’m getting this right, I need the following…
1080p@960fps 1/960 shutter
4k@3840fps 1/3840 shutter or 1920fps with 1/1920 shutter
8k@7680fps or 3840fps with a shutter of 1/3840 to 1/8000th
and in maybe 10 years 16k@10000fps or 10khz with 1/10000th shutter

If i’m getting this correct, the ISO values would need to be what DSLRs shoot at, as high as (Guessing) 51200 or 102400.
And so this head cam would need a resolution of 16k@10000fps to match reality completely and at least 4k@1920fps or 3840fps to start off with…
I can interpolate 1920fps with a 1/1920 shutter from my new GoPro Hero 6, but need better interpolation software, and a computer to play this type of video on. I am intrigued by VR, but I’m more into video.

Like0Vote Down 
5 years 11 months ago
POVHFR Videos
Member
POVHFR Videos

Thanks so much!! I don’t have a 240hz display at the moment. I plan to get one in the near future with some more $$$.
I’ll try out the test of 60 > 120 > 240hz when I do get one.
This is the most detailed, educated response I’ve ever had on camera & display refresh rates, ever!
The holodeck startrek recorder sounds so fun! Maybe around the 2030s or something like that, we could achieve 32k “retina”resolution at 16,000fps or higher in the 3d recording or an infinite frame rate with enough innovation over the next 20 years. 16k should be out in the next decade perhaps, but at probably something more on the line of 1000fps – 4000fps at best (maybe).
But still this seriously changes my thinking on recording video and watching it back on a display at insane 5-digit numbered frame rates in the future.

And also I believe with proper technological advancements over the next three to five years we could see 2000fps@2000hz in 4k become high end for gaming, and cameras following suit with being able to record at least 4k at 2000fps being the size of a gopro perhaps (the only issue being battery life & storage, depending on storage innovation as well)

What I was also thinking about reading all of what you said (thanks! :D) is storage space. Even at h.265 compressing the dat of a h.264 video in about half maybe, we are only at about 4 – 16TB of affordable storage at the moment which is kind of a bottleneck for ultra high frame rate videos in excess of 1000fps at 1080p resolution, but it is still possible to see that kind of resolution in the next two years probably.

So definitely a 1080p video at 480fps on a camera and played back on a 480hz monitor would “look like real life” at the end of the day?

And a 4k video at 960fps perhaps according to the motion blur or pixel latency you mentioned above?

Like0Vote Down 
6 years 8 months ago
Acehavok
Member
Acehavok

Too bad I’ll have to sell my left nut for it.

Like0Vote Down 
6 years 8 months ago
Mark Rejhon
Author
Mark Rejhon

For the price of the kit format (barebones), you’d be surprised.

Like0Vote Down 
6 years 8 months ago
bennzl
Member
bennzl

Is it possible to order a barebones kit from somewhere?

Like0Vote Down 
6 years 8 months ago
Mark Rejhon
Author
Mark Rejhon

Link being announced before midnight ET today (before August 17, 2017)

Like0Vote Down 
6 years 8 months ago
Logun
Member
Logun

Is it a noticeable difference from 144 or even 240 (in your opinion)?

Like2Vote Down 
6 years 8 months ago
Mark Rejhon
Author
Mark Rejhon

Yes.

Although there are certainly diminishing returns, the difference between 240Hz (on this display) and 480Hz (on this display) is still visible to the human eye.

There is also a greatly reduced mousedropping effect.

Mouse effect during 60 Hz vs 120 Hz vs 240 Hz

Doubling the refresh rate, halves the distance between image-copies in the stroboscopic effect. This also applies to things like FPS games (e.g. flick-turns while staring at crosshairs — and then seeing the “stepping effect” of the turn).

In fact, even the 1000Hz report-rate of a gaming mouse is becoming a limiting factor — mouse manufacturers such as Razer, Logitech, SteelSeries, ASUS ROG, Zowie, and others, now need to release true-2000 Hz position report rates.

More info to come in article. Keep tuned.

Like1Vote Down 
6 years 8 months ago
RealNC
Member
RealNC

One would assume it’s 720p TN?

Like0Vote Down 
6 years 8 months ago
Mark Rejhon
Author
Mark Rejhon

4K 120Hz with reduced-resolution high-Hz modes.

Like0Vote Down 
6 years 8 months ago
POVHFR Videos
Member
POVHFR Videos

Seriously??? 4k120hz! Sweet!

Like0Vote Down 
6 years 8 months ago
open
Member
open

Nice so that would be 1080p at 480hz maybe. Really excited to learn more about this.

Like0Vote Down 
6 years 8 months ago
wpDiscuz

Recent Content

Blur Buster Forums

Blur Busters Forums

TestUFO Motion Tests

TestUFO Motion Tests