Thread: Cinematography of 2030s: Ultra HFR! I have witnessed realtime 1000fps on real 1000Hz.

Reply to Thread
Page 5 of 5 FirstFirst 12345
Results 41 to 50 of 50
  1. #41  
    Hello!

    NVIDIA partnered with Blur Busters (my business) to create a motion test for the new ASUS 360 Hz gaming monitor:
    https://www.nvidia.com/en-us/geforce...aming-monitor/

    PC Magazine says ASUS now has a road map to 1000 Hz (~2030s)
    https://www.pcmag.com/news/372914/i-...-a-better-game
    I met ASUS at CES 2020 and they confirmed they said that. A long term road map.

    I also met with a couple of HDMI Forum about the refresh rate race to future retina refresh rates. There is a need to go up a bit sharply on the diminishing curve of returns (144Hz->360Hz is a 2.5x jump much like 60Hz->144Hz is, while 240Hz->360Hz is toot tiny incremental). Eventually, retina refresh rates will merge with retina resolution -- there is now an engineering path!

    After that, there is a need to dramatically up the curve. 120Hz -> 480Hz, or 120Hz -> 1000Hz, to show hugely visible benefits in the diminishing curve of returns. Many articles say 240Hz vs 360Hz are tiny difference -- ignore those. Read those websites that write properly about the 360 Hz monitor, and understand how the diminishing curve of returns work.

    RED Cinema needs to pay attention to the 2030s road map -- we are a 2030s Road Map Business. Retina spatial resolution is almost fully milked, and it will be time to milk retina refresh rates. 4K was 5 figure priced back in the IBM T221 days. Now 4K is a Walmart $299 sale. 120 Hz is about to be commoditized with the upcoming iPhone and Galaxy. 240 Hz is still bleeding edge but will become more mainstream.
    Reply With Quote  
     

  2. #42  
    Moderator Phil Holland's Avatar
    Join Date
    Apr 2007
    Location
    Los Angeles
    Posts
    11,840
    Heh. I was standing right behind you at the Nvidia event at CES Mark. Thanks for that very useful tool.

    Camera-wise it will be interesting to see if 1000Hz ever becomes part of the equation. In gaming it will, though somewhere around 480Hz might end up being the end of the line. I agree fully that a higher refresh rate helps with the user experience in interfacing with programs and games.

    At the moment in our industry due to the stretching futures of common workflows the industry is mainly focused on 24, 30, 60, and 120fps content. I'm talking the live-action workflow here.

    And the most interesting thing surrounding cinematography and HFR is it's a highly debated subject. More so than Resolution or HDR. However, in HFR the potential to solve various motion issues related to higher contrast ratio HDR grades and display tech is there.

    Motion Cadence has a drastic effect on the emotional impact of a motion picture. Many are still exploring HFR's potential, but similarly some squeezed the lemon and got juice in their eyes.
    Phil Holland - Cinematographer - Los Angeles
    ________________________________
    phfx.com IMDB
    PHFX | tools

    2X RED Monstro 8K VV Bodies, 1X RED Komodo, and a lot of things to use with them.

    Data Sheets and Notes:
    Red Weapon/DSMC2
    Red Dragon
    Reply With Quote  
     

  3. #43  
    Senior Member
    Join Date
    Jan 2008
    Posts
    6,157
    Why the F am I subscribed to this thread, I don't even have a post here? It doesn't look familiar, and the original posts definitely should if I read them.
    An explorer explores new ideas and things, but somebody stuck in mud, complains.
    - But the explorer complains about how much better the territory he has discovered is instead.

    -Since when has denying that the possible is possible been reason? Such is the life of some skepticism.
    -Normalcy involves what is realistically possible, not just what you say.
    Reply With Quote  
     

  4. #44  
    Quote Originally Posted by Phil Holland View Post
    Heh. I was standing right behind you at the Nvidia event at CES Mark. Thanks for that very useful tool
    You are welcome. Yes, I partnered with NVIDIA to create that 360 Hz monitor test for both the NVIDIA CES 2020 booth and the ASUS CES 2020 booth!

    Quote Originally Posted by Phil Holland View Post
    Camera-wise it will be interesting to see if 1000Hz ever becomes part of the equation. In gaming it will, though somewhere around 480Hz might end up being the end of the line. I agree fully that a higher refresh rate helps with the user experience in interfacing with programs and games.
    Keep tuned. There is already ongoing business involving >480Hz. Such work won't be announced for a while yet, but rest assured, 480Hz is not the end of the line.

    Even 0.5ms MPRT versus 1.0ms MPRT is human-visible in the "See for yourself" ULMB test, see Pixel Response FAQ: GtG Versus MPRT. If you have an ULMB monitor, there's actually a way to configure your NVIDIA ULMB monitor to 0.25ms MPRT, 0.5ms MPRT and 1.0ms MPRT via the on-screen "ULMB Pulse Width" adustment.

    0.25ms strobe backlights will require sample-and-hold 4000fps at 4000Hz to match morion blur for a natural stroboscopic-free 360-degree shutter without impulsing.

    Even a 1/480sec SLR camera shutter versus 1/2000sec SLR camera shutter is still human visible when filming fast sports video.



    What most don't know is that camera persistence (shutter speed) and display persistence is additive. Sample-and-hold displays (majority of non-CRT displays) are a persistence of one refresh cycle. So 60fps camera video 1/60sec shutter, displayed on a sample-and-hold display, is 1/60sec + 1/60sec = 2/60sec of motion blur. So you've doubled motion blur. 1/30sec of motion blur for your 1/60sec shutter video.

    Source persistence = camera shutter
    Destination persistence = display persistence

    The motion blur mathematics are identical, which is why Blur Busters is such a big-time UltraHFR advocate.
    48fps HFR and 120fps HFR still has a nausea uncanny valley that disappears for a bigger percentage of humans when you brute-force your way out (low-persistence sample-and-hold).

    Even for 1/1000sec sports shutter is still hobbled by the motion blur of a 60Hz display, 1/1000sec (source persistence) + 1/60sec (destination persistence)

    The only way to achieve low-persistence simultaneously with sample-and-hold (no impulsing), requires consecutive frames at the target persistence, which means 1/1000sec source persistence and 1/1000sec destination persistence, for a resulting of 1/500sec persistence. Even that is still not the final frontier.

    480Hz Ultra HFR will still be somewhat hobbiled still, e.g. 1/480sec creates 16 pixels of motion blur at 7680 pixels/sec panning (one screenwidth per second panning) at 8K. Thus, 8K 480Hz will still have motionblur limitations that will still require pushing the retina refresh rate curve far well beyond 1000Hz.

    Now, lest someone cite the famous Fighter Pilot 300-500Hz Study -- remember that's a non-lumen-compensated momentary-visibility test, rather than a stroboscopic visbility test or a motion blur test. It did not test other variables that creates differences between displays and real world.

    At the moment in our industry due to the stretching futures of common workflows the industry is mainly focused on 24, 30, 60, and 120fps content. I'm talking the live-action workflow here.

    And the most interesting thing surrounding cinematography and HFR is it's a highly debated subject. More so than Resolution or HDR. However, in HFR the potential to solve various motion issues related to higher contrast ratio HDR grades and display tech is there.

    Motion Cadence has a drastic effect on the emotional impact of a motion picture. Many are still exploring HFR's potential, but similarly some squeezed the lemon and got juice in their eyes.[/QUOTE]
    Last edited by Mark Rejhon; 10-17-2020 at 04:36 PM.
    Reply With Quote  
     

  5. #45  
    Big News: Christie Announces 480 Hz Capable Cinema Projector

    Check out the news: Christie Announces 480 Hz Capable Cinema Projector for Ultra HFR: Griffyn 4K32-RGB Laser Projector.

    While 480 Hz Ultra HFR on the new Christie projector is limited to 1080p, and DLP has some difficulties with 24-bit-or-better temporal dithering at high Hz -- the three-chip nature and the lower resolution (spatial dithering via scaling for better color depth per ultra high Hz pixel), hugely compensates. 480Hz on this Christie looks better than 120Hz on a consumer 1-chip DLP. Ultra high Hz DLP has formerly often been subject to color depth reductions or banding/contouring issues (noisy/dithered/rainbow motion blurs, rather than a nice SLR-style phoographic motion blur), especially in consumer DLPs.

    The first Ultra HFR tests on Christies will help convince the world of the legitimacy of Ultra HFR, and even 480 Hz is currently in the plans/roadmap as not the final frontier (1000Hz+ is in active development)

    Blur Busters Ultra HFR FAQ Updated

    Additional information has now been added to the Blur Busters Ultra HFR FAQ containing more information about the additiveness of source persistence (shutter) and destination persistence (sample-and-hold motion blur). Motionblur has been a big problem since we discontinued projector strobing & CRT impulsing, and standardized the world around sample-and-hold, even in electronic cinema projectors.

    But we can still brute-force our way out of that, thanks to upcoming new ultra-high-Hz displays. ASUS now already has a road map of 1000Hz displays in the next 10 years. Simulating 1ms motion blur requires 1000 consecutive 1ms frames.

    Ultra HFR has a bright future with some of us already, even if other cinematographers may not believe in it yet (due to the nausea of 48fps and 120fps HFR, when UltraHFR just leaps past the nausea uncanny valley...).

    It did not take until recent tests and virtual reality to discover that 48fps and 120fps HFR is still in a nausea uncanny valley -- a researcher needs to test 24fps thru 1000fps on a true sample-and-hold 4K 1000Hz display, once those become available. Least nausea occurs at 24fps and at max Hz, with a hump of nausea at intermediate rates that still creates blur on sample-and-hold displays.

    This is so new a discovery, few academic papers don't even exist yet, mainly discovered by commerecial businesses in VR and other industries. But should be a new display research thanks to full-color-depth Ultra HFR developments finally becoming possible with fewer weak links (i.e. subject to the Vicious Cycle Effect problem. This mainly affects wide-FOV retina-resolution displays like VR and cinema-sized screens, as well as ride-simulator screens. This is where Hz limitation is so massively amplified, that even 480Hz is a limiting factor. It prevents the display from feeling like a window to real life, or needs a strobing bandaid (CRT emulation, like most VR headsets need, due to finite Hz limitation).
    Last edited by Mark Rejhon; 06-19-2020 at 01:11 PM.
    Reply With Quote  
     

  6. #46  
    Quote Originally Posted by Phil Holland View Post
    In gaming it will, though somewhere around 480Hz might end up being the end of the line. I agree fully that a higher refresh rate helps with the user experience in interfacing with programs and games.
    Famous last words -- 640K limit is end of line, 1080p is end of line, etc.

    HFR has a nausea valley that needs a big jump to bypass. The smart ones that have seen UltraHFR in quadruple-digits agree that triple-digits are still too limiting to perfectly match real life, thanks to remaining whac-a-mole effects (even if flicker is gone, there's still stroboscopics and motion blur).

    I'm a Fan of Filmmaker Mode 24fps.
    But Need 1000fps To Bypass Nausea Uncanny Valley of 48fps & 120fps HFR


    I am a big fan of 24fps. I just personally adore Hollywood Film Maker Mode. But anyone who genuinely saw experimential 1000fps Ultra HFR is superior to 48fps and 120fps in HFR comfort, [i]if HFR must be used[i]. Caveat of scientific tests: At full resolution, please see mandatory scientific requirements for successful Ultra HFR.

    The great news is that there's an engineering path. ASUS publicly confirmed they have roadmapped to 1000Hz, as told to me directly, and also quoted in PC Magazine ("road to 1000Hz").

    As I've said in earlier posts, 4K was a $10,000 curousity (IBM T-221) twenty years ago, but today it's a $299 Walmart special. Ultra HFR isn't going to be unobtainium forever.

    Quote Originally Posted by Phil Holland View Post
    At the moment in our industry due to the stretching futures of common workflows the industry is mainly focused on 24, 30, 60, and 120fps content. I'm talking the live-action workflow here.
    Yep. First things first. These are 2020s HFR rates.

    However, we're the equivalent of 1980s Japense MUSE HD researchers, or early 1990s 4K researchers. We're already working on Ultra HFR, and already working on >480Hz as an actual business.

    This will be relevant to cinematographers coming in 2030s for wide-FOV situations (VR, simulators, ride simulators, speciality digital cinema, fast-motion live action, etc) where the Vicious Cycle Effect is most horrendously amplified. The material where you want to use a 1/1000sec (or faster) SLR camera shutter, is the precise situations where you want destination persistence to at least match, for reality emulation use cases. At these leagues, 480Hz is not even enough for a wide-FOV retina-resolution display to pass a reality turning test (display indistinguishable to window out of a fast-moving car or train).

    It won't matter for desktop displays as much, nor small TVs viewed across the room, because you need more pixels over longer eye-tracking time (aka retina-resolution wide-FOV displays) which is why Ultra HFR will hugely benefit IMAX, VR, simulators, of the 2030s. There are too many weak links preventing all six bullets from being satisfied, but we're slowly whac-a-moling them out of the way.

    Quote Originally Posted by Phil Holland View Post
    The most interesting thing surrounding cinematography and HFR is it's a highly debated subject. More so than Resolution or HDR. However, in HFR the potential to solve various motion issues related to higher contrast ratio HDR grades and display tech is there.
    Yep.

    Blur Busters role is to mythbust all those descendants of "Humans Can't See 30fps Versus 60fps" Or "That Fighter Pilot Test Proves 500fps Is Limit" (Wrong: It's a non-lumen-compensated momentary-visibility test; it did not test motion blur effects or stroboscopic effects). We're the role of Hz mythbusting, which pays the Blur Busters bills, obviously -- as long as people disbelieve the metaophorical Hz-equivalent of Einstein and stick to only pre-Einstein Newtonian math, we'll be in the business for a long time -- vision is a highly complex subject.

    Quote Originally Posted by Phil Holland View Post
    Motion Cadence has a drastic effect on the emotional impact of a motion picture. Many are still exploring HFR's potential, but similarly some squeezed the lemon and got juice in their eyes.
    Yep.

    There's a huge HFR Nausea Uncanny Valley that will require jumping to quad-digits.

    Very few people have seen this for themselves -- for a proper scientific UltraHFR test that is aware of the Vicious Cycle Effect (higher-Hz wider-FOV retina-resolution simultaneous = amplify Hz limitations of sample-and-hold).

    As a 2020s decade stopgap, impulse-driven 120fps 120Hz HFR is a great stopgap, since 120Hz is above flicker fusion thresholds (although still hugely prone to stroboscopic effects of finite frame rates). Problem is not all displays impulse-drive (strobing, phosphor, black frame insertion, etc) to allow 1ms flickers of each frame, and DLP will lose color depth/brightness when that is done. However, 8K 120Hz impulsed 1ms looks just like a beautiful CRT -- very pleasing to look at (as seen in low-persistence impulsed virtual reality, which Blur Busters helped). Certainly 120fps HFR is not the final frontier.

    I'm part of an NVIDIA research paper (Page 2), of Temporally Dense Ray Tracing (240Hz). Open to collaborating with other researchers on Ultra HFR papers, since it's such a new study that was discovered by VR/gaming long before university researchers. The HFR Uncanny Valley effect is somewhat of a subject that is a little understudied, and it will help cinematographers understand Right Tool For Right Job.
    Last edited by Mark Rejhon; 06-19-2020 at 01:12 PM.
    Reply With Quote  
     

  7. #47  
    In September, Samsung television researchers just used my Blur Busters work for new approaches to measuring motion metrics: 57‐2: A New Approach to Motion Frequency Metrics Quantifies Motion‐induced Blur, with my TestUFO mentioned in citation number 9.

    Myself and my sites (Blur Busters, TestUFO) are now mentioned in over over twenty different research papers now, as I am such an ideas-trailblazer in this territory.

    Things that are common-sense to me that I researched years ago are now fully confirmed in full scientific tests by display researchers; and being utilized to advance related arts as we speak. (Motion blur, displays, etc)
    Reply With Quote  
     

  8. #48  
    Senior Member Karim D. Ghantous's Avatar
    Join Date
    Oct 2011
    Location
    Melbourne AU
    Posts
    1,966
    Quote Originally Posted by Mark Rejhon View Post
    Things that are common-sense to me that I researched years ago are now fully confirmed in full scientific tests by display researchers; and being utilized to advance related arts as we speak. (Motion blur, displays, etc)
    I kind of know how you feel - sort of. When you see what the future will bring, more or less, the people who took part in the previous revolution (digital photography) think that nothing will change after it. "Mirrorless cameras? Nah, they're just toys. Adobe Flash will become extinct? Nah, HTML can't replace it. 4K cinema cameras? Nah, story matters, resolution doesn't. The iPhone will lead the market? Nah, Microsoft sets the standard. CGI will replace photography? Nah, it's not real enough." Etc, etc, etc.
    Good production values may not be noticed. Bad production values will be.
    Pinterest
    | Flickr | Instagram | 1961 (blog)
    Reply With Quote  
     

  9. #49  
    Senior Member Christoffer Glans's Avatar
    Join Date
    Jun 2007
    Location
    Stockholm, Sweden
    Posts
    4,369
    I've always been fascinated with how HFR gets marketed as being closer to human perception. Aren't our eyes and visual cortex adapting framerates based on the Hz of the light sources? If you stand in a room with fluorescent lights that have a certain Hz value, you will see that your vision adapts to that Hz in motion. If you watch motion in daylight, that's not the same as watching HFR even at 120 Hz, because at 120 Hz your eyes adapt to that type of motion update speed. I don't think that 48, 60 or 120 Hz feel unnatural to watch because we are used to 24 fps cinema, I think it feels unnatural to us because it's even more unnatural compared to real-world motion. In daylight, there's a certain quality to motion blur that 24-30 fps is closer to replicate than HFR. In daylight or natural light sources, there's no Hz update value to adapt to, so the motion blur is very specific to our vision. That motion blur is non-existing in HFR videos. The sharpness of HFR videos creates an unnatural way of viewing motion that I believe we've mistaken for being closer to human perception when in reality it might be the opposite.

    For HFR to work, it has to replicate how humans perceive motion blur under natural light sources. Thinking that Hz update values are the only thing needed to replicate a human perception of motion is excluding a lot of other things related to how we perceive motion.
    "Using any digital cinema camera today is like sending your 35mm rolls to a standard lab. -Using a Red is like owning a dark room."
    Red Weapon 6K #00600
    Reply With Quote  
     

  10. #50  
    Senior Member Eric Z's Avatar
    Join Date
    Dec 2010
    Location
    Israel
    Posts
    534
    The way I see it - For humans, the perfect response time, or refresh rate, is 1000 Hz (1ms).
    That's where we can stop.
    Digital pens for phones and tablets should respond at 1ms for us to perceive the response as real-time and fluid, without delay. Currently we're at 9ms (iPad Pro, Galaxy Note 20, etc...).
    Same goes for screens/monitors - response time and refresh-rate should be 1000Hz. Slowly but surely, we're getting there. Currently we're at 360Hz (Asus PG259QN, Acer Predator X25, MSI Oculux NXG253R, etc...).
    "Don't under-estimate the amateur photographer. The latest HW & SW empower him to achieve results just as good as yours".
    Nothing beats a Master Prime!
    Gigabyte Z77X-UD5H, 24GB RAM, i5-3570K, Intel 520 120GB SSD - Win7 Pro SP1 x64, Dell U2713H 27" 10bit @ 2560x1440@60Hz (via miniDP).
    Reply With Quote  
     

Tags for this Thread

View Tag Cloud

Posting Permissions
  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts