What is the Frame Rate of the Human Eye?

Eventually the frame rate discussion turns to what the human eye can see. And that conversation is full of myths and misunderstandings. For those daring few people who read the description blurb, here it is, the human eye HAS NO FRAME RATE!!! That's it, you can stop watching the video now.
HOWEVER...
There are some interesting biological phenomenon that can be categorized as "frame rate of the human eye" if you're willing to abuse the terminology. And one of those numbers is 10 Frames Per Second!
Check out Wired's Video on Wagon Wheel Effect with Dr. Eagleman for a longer discussion:
• Why Your Brain Thinks ...
#FrameRate, #Biology #

[@Undy1]

I like and agree with your conclusion - The human eye HAS NO FRAME RATE. I think debating how many frames the human eye can see is counter productive because that's not how eyes work. Instead a better question to ask is - what FPS is needed to completely fool our eyes into 'thinking' that they're looking at reality? Now that's a completely different story. Take the cursor on a screen example - at 60Hz we saw 2.5 discrete cursors, at 144Hz we saw 6. But if the cursor was a real object neither our eye or a camera would see multiple discrete cursors - instead we would see a continuous smear. To simulate that effect with a screen we would need a refresh rate so high, that the moving cursor would still overlap a significant amount between refreshes. Again, we're looking for a way to fool the eyes into 'thinking' they're looking at real moving object instead of apparent motion. Though the cursor is a very extreme example and would likely need at least 1000Hz refresh rate to appear continuous (this strongly depends on the relative size of the cursor and it's speed - the smaller and faster it is, the more Hz you need to make it appear continuous instead of broken into a 'phantom array'. Now for a less extreme example - take a standard first person shooter - you have a much more natural motion of objects in a scene - moving people, you moving and turning yourself etc. Sadly it's an example you did not include in your analysis - when presented with the sentence "I can tell the difference between 60Hz display and 144Hz display" you immediately jumped into the cursor explanation, completely dismissing the fact that people CAN tell the difference between these refresh rates IN-GAME. And yes, I am one of those people. Even when you're not looking for it, 144Hz is just so much more fluid and natural in-game than 60Hz. It is also so much easier to track fast moving objects at higher refresh rates. For example in first person shooters you obviously have to shoot your opponents who are moving - and that depends on your ability to track the moving object and keep pointing your crosshair at it - at higher refresh rates this is much easier because all the motion is more fluid and you get more frequent updates which allow you to make faster and more precise adjustments in real time (watch the LinusTechTips "Does High FPS make you a better gamer? Ft. Shroud" for a very detailed analysis of this). And that wouldn't be possible if our eyes were running a discrete framerate or if they weren't able to perceive more than say 60FPS as some people are claiming. If you had a camera at a fixed framerate, then increasing the refresh rate of the monitor that the camera was looking at wouldn't give the camera an advantage - the camera wouldn't get more frequent updates than it's own internal framerate. Our eyes do get the advantage of 144Hz over 60. Now the next step - 240Hz is smoother than 144Hz but the difference is not as staggering as between 144Hz and 60Hz. Still some people can tell the difference and 240Hz is still not enough to completely fool our eyes in every single case. For example you can still see the phantom array while moving the cursor on the desktop. 240Hz also wouldn't be enough to simulate the J. F. Schouten experiment on-screen and here's the math to prove it and calculate the minimum framerates to do so: Let's take a disc divided into 15 equal sectors, each 24° wide. Let's say that for the effect to occur we need the sectors overlap by at least half of their width between frames so 12° Alpha occurs at 8-12 cycles (revolutions per second). Beta: 30-35 cycles Gamma: 40-100 cycles To have the sectors overlap by half of the width of a sector we need there to be twice as many frames per revolution as there are sectors - 15x2= 30 Alpha at 8 cycles * 30 frames per cycle = 240Hz, 12 cycles * 30 frames = 360Hz Beta: 30 * 30 = 900Hz, 35 * 30 = 1050Hz Gamma: 40 * 30 = 1200Hz, 100 * 30 = 3000Hz Keep in mind that this is only to overlap the sectors on the fan by only half their width each frame. If we wanted them to overlap by 3/4 we'd need to multiply all the refresh rates by a factor of 2. I'd say that would probably be enough to fool our eyes into seeing the exact same effect on screen as we do in real life where the motion is completely continuous instead of broken into frames. But this is, just like the cursor - a very extreme example - in most cases of everyday stuff, much lower refreshrate would suffice to fool our eyes. Let's take a 3m long car going by us 30m/s - let's say we need it to overlap 5/6 each frame - it takes 0.1s for the car go past our eyes in it's entirety and so with the required overlap we need to have a frame time of 0.0166s which translates into 60FPS. But if we were to replace that big car, with a 30cm long drone going the same speed and same overlap requirement we would need 10 times the framerate or 600FPS for it to appear continuous instead of breaking into a 'phantom array' (several discrete images of the drone visible at once at different points). And we know that in real life things don't break into phantom arrays and everything appears continuous no matter the speed. That is of course because The human eye HAS NO FRAME RATE, instead it sees in a more 'continuous' way. So in conclusion, the human eye HAS NO FRAME RATE but if we wanted to fool our eyes into seeing things on screen the same way we do see reality then 600FPS would be enough for the vast majority of everyday stuff, but there are some edge cases which would require 3000Hz or possibly even more.

[@FilmmakerIQ]

That's also the kind of math necessary to understand how to fool a camera shooting a discrete frame rate. ;) To me, the eye question is far more interesting. Because it also provides insight into "hacks" like strobing impulse displays.

[@UncleMarty]

Agree about the 60fps/Hz comment. That seems to be the sweet spot for most people.

[@shanroxalot5354]

@Filmmaker IQ I love that hack, these super bright HDR displays can use their brightness budget to dip more heavily into blochs law, and give superb motion clarity per refresh. Though the phantom array effect is still an issue.

[@s.bgaming6134]

so minecraft on a good gaming computer is reality ok.

[@rahmani144]

This was PERFECT... THANK YOU!

[@dnoordink]

In other words, we don't see reality. Our brain makes it all up!

[@OmarIsmaili-ep9hh]

not really

[@NormansWorldMovies]

Close, but what we see is a recreation of reality by our brains.

[@douglasboyle6544]

Once again, you killed it with the depth and the clarity. I gave you a standing ovation.

[@benwaardenburg]

Your explanation at 10:40 kinda blew my mind. I had to rapidly play/pause to see both images side by side and you're totally right, the one frame of the candle compared to the three frame have different perceived brightness. Super cool!

[@Rilumai]

A little tip, the < and > keys will move the video one frame at a time when paused. Much easier than trying to time the pause, haha!

[@GregConquest]

@@Rilumai Thank you for pointing this out, Rilumai. I've often found myself struggling to isolate particular frames on youtube, for example on explosions, crashes, magic tricks, etc. And now it is made much easier. Just pause the video near the moment, and then use the < and > keys to move frame by frame. On VLC Media Player*, FWIW, the E key advances one frame (many other keyboard shortcuts for features of VLC not found elsewhere are listed in Preferences - Hotkeys; you can even change the hotkeys and their functions there). * If you don't have VLC on your Mac, Windows, or Linux computer, give it a try. It is open-source and great!

[@Rilumai]

@@GregConquest Hey, no problem! And I love VLC, I use it all the time.

[@serraxer]

Ofc it will be brighter cos you persieve it longer, so until you percieve difference in brightness you are a human who can percieve faster fps lets say 60 120 240 fps. So yeah your eyes definetly can handle High framerates films not just 24 fps.

[@FilmmakerIQ]

​​@@serraxer if it's brighter because you perceive it longer that's more proof that it's closer to 10fps. That's how exposure works... The longer you let it expose the brighter it is. Hence 10fps. But we don't actually see a frame rate.

[@arbabraiyan8200]

welcome back john, always loved your in depth explanations on film topics

[@joshzaring]

thanks for spending the time to make this AND for the research involved.

[@UlrichBeinert]

This is such a fascinating and captivating video. I come from photography (specifically, originally astrophotography, where specialized CCD cameras are used) and I remember someone from that field telling me the eye has an "integration time" of 1/10 second. Not a frame rate, but rather in all that fluidity of the organic, biological system that is our visual apparatus, a "shutter speed" of 100 ms, so to say. Excellently confirmed by your experiment with the candle! BTW, I do see flicker on a lot of LED lights where others don't. I always thought I'm crazy til I met others who could also discern it (particularly in peripheral vision - thanks for confirming this, too!). It's hard to find good dimmed LEDs where I don't see the flicker!

[@andresep]

This is the best explanation I've seen on KZbin about the phenomenon of human vision!

[@jefffaddis8169]

This channel is pure awesomeness! Thank you for all the entertaining information!!

[@richjames2540]

Great video, mature and informed. Having spent many years editing both on film and video , there is one topic that I would like to add to your excellent piece. There is a good reason why the eye has no frame rate which is that we do not see the light that the eye gathers. Human vision works on the brain taking input from the eyes and trying to find a match that is updated if it does not see what the brain hypothesized. This is why you can use a wipe cut without viewers noticing it, why people recalling events that just occurred think they saw what they did not, slight of hand by magicians and the problem of flicker from 50Hz CRT's. I used to suffer terrible migraines from 625/50 crt's until I discovered having daylight through a window in the Edit Suite cured the issue. So cool I found your content, keep up the good work.

[@FilmmakerIQ]

Yup, there is so much processing that the eye does before any signal is sent to the brain. It's not like the optical nerve is sending a picture, it's sending some visual sensation... What people don't realize is the BRAIN is responsible for the vast majority of human vision, not the eyes.

[@909sickle]

An interesting nuance, if I show you an animation at slower rates (e.g. 3, 1, 0.5, etc fps), you can still see movement if you choose to. So, 10-12fps is not where we suddenly see continuous movement, it's where it becomes difficult to see individual frames. Subtle, but very meaningful distinction. As you say, we don't seem to see in quantized frames, so the 10fps seems more analogous to shutter speed. In reality, the limitation is probably either neural network speed (how fast your brain can process the data before the next frame arrives) or memory (how much mental space you allocate per unit of time, forcing your brain to overwrite information from the previous frames with information from new frames). You could probably test which one it is by finding your threshold for apparent continuous motion (the slowest frame rate that looks smooth) and then increasing the brightness. If the footage starts to stutter at the same frame rate, the bottleneck is probably NN speed (the extra light allowed your framebuffer to fill up faster, giving you more time to see independent frames). However, if you make the animation smaller or less detailed and it appears to stutter at the same frame rate (but at the same time does not stutter with more brightness), then the bottleneck is probably memory (mental "framebuffer" size). And it's not completely either/or. Alleviating one bottleneck can cause the bottleneck to move to the other side. In nature, things tend to be balanced, so I'm guessing the bottleneck is pretty close on both sides. I'm just making this up now, so my test logic could be flawed. Also, the algorithm might not be main culprit. Frame rate might just be really interesting. YT regularly recommends to me your videos and live streams. The frame rate videos are always very compelling to click on. It's also because anyone who knows you, knows there's drama and a story around you and frame. So we know there's going to be extra emotion and passion in the video.

[@FilmmakerIQ]

Be extremely careful making analogies like that. The brain and visual system are not cameras or traditional computers. There isn't memory in the sense of an accurate record of stimuli. As you go further down the road of this stuff you start to realize that ultimately visual perception is largely a function of what we THINK we saw. It's all subjective.

[@909sickle]

@@FilmmakerIQ I agree. By neural network I mean actual networks of neurons which computer NNs were named and modeled after.

[@FilmmakerIQ]

Still I think biological neurons and computer neural networks aren't as similar as we think they are. Just because they share a common word doesn't mean they are the same thing.

[@CodeKujo]

@@FilmmakerIQ Especially in the visual system, where so much signal processing has happened before it even reaches the brain.

[@HURSAs]

It's not an accident that the graphics in the Bloch's Law were having a logarithmic mathematical expression for log T (Time duration) and for the log L ( Luminance ) - both factors have cumulative effects over Time and the logarithmic function describes perfectly the asymptotical limiting behaviour of the biological visual sensory system - in Neural Network science both Memory (data as a signal or noise accumulation) and Time (number of computational cycles and their duration depending on the complexity of the apparatus) are of signifficant importance!

[@matheus5230]

I remember, when I was a kid and teenager, that shooting a TV with my phone camera always looked weird, there was always some darker horizontal bar scrolling across the image. I guess that must have been because of the flicker, and modern TVs have far less flicker. When you talked about being able to perceive a fast flashing light, I thought you were gonna bring up illusions like if you flash a red light very quickly, you see it's red. But if you also flash another light of another color, like yellow, quickly enough, you will perceive a color that is the mix of red and yellow. Great video, so cool and interesting!

[@Quetzalcoatl0]

Can't agree more! Some of the experiments are really fascinating.

[@rafograph4714]

Such detailed info as always. Thanks John!

[@VanCamelCat]

Dude. The bottom line conclusion of this video was an art of a speech 👏 Beautiful.

[@MrBerniemcgovern]

Fantastic presentation. This helps clear things up like crazy, thank you.

[@tcallaha]

Fantastic explanation! This should be required watching for anybody working with motion-based media.

[@alex_montoya]

One of the very best channels on the internet. Kudos!

[@-MertArda]

An actual well-studied video, really good job!

[@SrChalice]

Define well studied? This is nonsense

[@FilmmakerIQ]

Morons will always think actual science is nonsense.

[@SaturnCanuck]

That was great John. So loves these essays.

[@rizzo-films]

I love this work that you did. I'm a cinematographer, so I know that if I try to shoot at 1/10 of a second, my video is going to have so much motion blur and it will be choppy because it's lower than the frame rate. But the reason my eyes don't see choppiness IRL is because they don't have a frame rate! But they DO see the motion blur. It's just so smooth because it doesn't have a frame rate that it needs to sync up with. It reminds us that we might be forcing the technology vs. biology comparison too much as a trivial attempt to understand our own biology, losing out on some other points of knowledge inherent in the biology. Maybe one day our screens will be more like the human eye? No more refresh rate or frame rate, just a persistent image flowing with the arrow of time instead of rigid increments of time. Similar to how a film projector is the inverse of a film camera, I'm thinking of a monitor that is the inverse of a human eye. Right now, monitors are more like the inverse of camera sensors, not eyeballs. Maybe it would need to mimic biology more? Some people equate rods and cones to pixels, but pixels are arranged in a fixed 2 dimensional grid. What if the screen was actually a 3 dimensional vat of organic "pixels" floating around in a malleable state rather than fixed? Maybe the next evolution of OLED screens. Sort of like the grain in video noise but they're actual pixels, dancing around extremely fast to generate a fluid, persistent image, and the pixels would move along 3 axes instead of 2. It would change the approach to pixel resolution, similar to the human eye. Take that a little further with machine learning and maybe... MAYBE... we can have something like "synthetic subjectivity", where each time we viewed something it was interpreted in a slightly different way by the display itself, sort of the inverse of the way the brain takes in and interprets image data subjectively, and you can toggle it on or off on the TV lol. But maybe that's taking it too far? It would at least be interesting and perhaps a new form of storytelling, like Rashomon but if Kurosawa was actively showing us a slightly new perspective every time we watched the film. We'd not just be viewing a pre-recorded thing, but in some sense the thoughts and memories of the machine, what it's "seeing" with its mind's eye and how its memory recall process evolves that story. I THINK this would be a more accurate comparison to human vision than current cameras and monitors.

[@leoguado2]

Love this, thank you for the explanation!

[@zsejuestudios]

Awesome video, I do stop motion and I have always asked this question. Thanks for answering

[@metarumashin6045]

Great video, long time ago I wondered how much resolution the human eye had and the answer was similar, it's too complex to put a single number, but never thought about the frame rate, definitely it's impossible to measure the eye with camera technology terms

[@downvotesdipsdownvotemeb3322]

What a refreshing video! Very well done, loved it

[@RendezvouDoo]

Love your content!

[@morpheus220276]

Thank you for your work. 😊

[@aarongrooves]

Fascinating!

[@Account.for.Comment]

Great video again. Eventhough, I don' t care as much for framerate arguments. Whomever get to be director of photography or any of his/her superior, can decide whether or not to trust, experiment or rejects of the hype in new fads in the field. 24fps will stay, not because one youtuber like it, but because the industry like it, and that already explained before. One curiousity for me is how the framerate, motion blur affect the pacing of the film/video from the view point of the audience. Don' t know how to explain. Whenever I watch movies or videos with higher frame rate, I felt that my attention got stuck a bit longer on the scenes being shown, instead of the usual. However, they also feel a nauseating feeling. It almost like as Plinkett put in "don' t notice it but brain did". With the candle frame comparison, I've had a better idea now.

[@MrRoboticBrain]

This is definitely the best Video on this topic i have seen yet! Including the properly addressed epilepsy warning, which too many people don't care about or purposefully screw up! The only information that's missing for me is: "What is the minimum frame rate to make the phantom array effect imperceptible"? This bothers me because some LED lights and projectors have this weird effect where the colors separate or turn into discrete dots and dashes when you move your eyeball to another point in space, thereby creating the phantom array effect. This discussion leads me to believe that this might be actually impossible because "The Eye has no frame rate" and only relies on exposure and persistence of vision. This basically means unless your retina is continuously exposed by the light - creating blurred motion - you WILL get a phantom array! Therefore the critical frame rate depends on distance and angular speed because the phantom array effect is directly coupled to the resolution limit of the retina and the circle of confusion of the eye's lens. A very fascinating conclusion! A small interesting Note: in the example with the candle; the second version with the 3 and 6 frames, the left image still looked a bit darker for me. I guess this only drives home that biological vision depends on so many factors. Like eyesight, alertness, fatigue etc...

[@FilmmakerIQ]

To get rid of phantom array effect, you've got be reaching 1000hz and above - as high as maybe 20,000hz depending on the speed of the motion. Basically it's a function of human visual acuity - what's the smallest detail you can see... now you need a refresh rate so high that each new frame advances by that smallest detail. Then the phantom array will look like what it would in real life - a blur.

[@babaarcuszatir]

Love these videos, I’ve been waiting for one like this so long! 😀 Learned a lot from the previous ones, thank you for all the educational resources! 🎉

[@franosbornblaschke3694]

Such a good Channel John! : )

[@MortyMizrahi]

This might be one of my favorite videos in a LONG time. I will have to rewatch it a few times bc so many nuggets information . Absolutely love it!!

[@lapq3590]

It might have been on purpose, but I think most of this video isn't having the same discussion as the general public. Maybe the Critical Flicker Fusion is actually the point, though. As far as I know, what people mean when they talk about the "framerate of human eye" is at which video framerate threshold the human eye stops being able to spot any practical difference in fluidity in apparent motion. But I'm glad you wandered through so many aspects of this topic. I knew none of the information you brought to the table, and it was all very interesting. I thank you for this excelent video, sir.

[@FilmmakerIQ]

The problem the "general public" as you described it has is they don't understand the basics of vision and the human eye. You want to answer the question what is the threshold of fluidity of apparent motion without first getting a firm grasp on the biology of the eye. But I answer your proposed question in the follow up video. The problem the "general public" has is the answer to that question doesn't involve the frame rate of the capture device - in fact the frame rate of the eye is entirely irrelevant to the discussion of what's the highest frame rate you can see- instead its about acuity: kzbin.info/www/bejne/bavOqmCporeSkLs

[@smaraggi]

Great video! How to make from an apparently very simple matter, very interesting observations.

[@d4v01d]

Good video. I came here after seeing some digital cameras site articles, based on this video, drew entirely the wrong conclusions. Interestingly, in the candle test I could very easily tell the difference in brightness in the extended version. I can also detect strobing in some led light bulbs. Not so much with fluorescent bulbs though .

[@FilmmakerIQ]

Bloch's law isn't a constant, like all things psychophysics it's a range.

[@samp.8099]

The frame-rate man strikes again!

[@user936]

it's the sequel to the 24 series no one expected 😁

[@billyhess5263]

When you have a last name like “Hess” it’s in your blood to be a nazi about something…look at Rudolph Hess, worlds first grammar nazi…frame rate nazi????

[@FilmmakerIQ]

30FPS??? RIGHT TO JAIL. RIGHT AWAY

[@kamathln]

The Anti-FrameRate man!

[@drajitshekher]

Really liked your presentation. But I am going to have to save it to watch a few more times to unpack everything

[@8lec_R]

What an amazing video. Learnt so much

[@lamcho00]

You really dove deep for the preparation of this video. Consider me impressed! I do agree the human perception of movement doesn't have a frame rate. As far as I remember, your initial argument was we shouldn't strive for more frame rate in movies, because 24 is all you need for movement and more than that is just extra effort when creating effects and filming a movie. I still disagree with that statement. Movement is still more fluid at higher FPS, so there is room for improvement. Also 3D movies at 24 FPS stutter when there are fast moving 3D effects. The stutter is gone on 48 FPS projections. Now I can't say if the increased production cost makes sense just for more fluid movement. I can agree that probably that's the deciding factor. By the way, in the experiment you do at 5:44 with the 2 monitors and the cursor shadow, you should distinguish between "signal refresh rate" and "LCD monitor matrix response time" and "screen backlight refresh rate". Most monitor manufacturers market their "LCD matrix response time" by citing "Gray-to-Gray" value of 1 ms. But in reality their "Black-to-White-to-Black" response rate is much slower. That's the reason you see multiple cursors with the naked eye and on a 24 fps camera. On top of most of the older LCD monitors used fluorescent backlight which flickers with power supply frequency at 50/60Hz. Newer gaming monitors use white LED backlight, but control monitor brightness by strobing the LEDs. So you can have a 144 Hz monitor but configured at low brightness you'll still see something like 60 FPS. Gamers take advantage of this backlight strobing to minimize those shadows caused by LCD/pixel response time, by synchronizing the signal refresh rate with the backlight strobing, so when the frame changes, the backlight is off for a millisecond or two. This lowers your effective monitor refresh rate, but makes moving objects with cleaner edges and easier to identify. You might want to check the BlurBusters web site for a better explanation. The same effect was also visible on the older CRT monitors, but the fluorescent layer (hit by the electrons) would shine for a much shorter amount of time. I still miss my old CRT monitor. For me personally, I stop distinguishing motion fluidity difference at around 80 FPS. So if fast action scenes would run at higher FPS in the movies, it would be a great change, at least for me.

[@FilmmakerIQ]

You have my argument for why 24 fps is here completely wrong. But in one of the next videos I'm going to talk about what I think is an underappreciated feature for watching movies and it definitely relates to backlight: BFI and flicker. And FYI I discovered Bloch's Law from Blurbusters! :)

[@sk8erbyern]

Camera pans are the most obvious culprits in 24 fps cinema. They honestly look horrible, no exceptions. Maybe some shots/scenes where the camera is stationary but the objects are moving look better at 24 but the moment camera itself starts a rotation it looks horrible at 24. As a gamer after getting a 165 Hz screen I honestly cannot even stand 60 anymore so I have an idea where this guy is coming from. I ridiculed 30 fps for over a decade and advocated for 60. Of you spend your whole life admiring masterpieces at 24 without experiencing higher framerates them ignorance is bliss

[@CodeKujo]

cool, so nice to have the term flicker fusion to describe my experience, and neat that good low-light vision (which I also have) is related. A high framerate crt driven at 60Hz, or half-wave rectified LED strings are noticeably flickery to me, even without “motion” (though certainly even more obvious as my eyes saccade)

[@FilmmakerIQ]

You're the first person to use the word cicade! Kudos. That's a topic I didn't even get into.

[@CodeKujo]

@@FilmmakerIQ :D (Just looked up how it's spelled now; sorry I got it wrong before.) You already covered phantom array well, and had so much to cover in this video!

[@FilmmakerIQ]

When I was shooting that spinning pattern test. I could essentially get the spinning pattern to freeze by flicking my eyes about. So it's really into saccades adds another layer of complexity!

[@EdwinvandenAkker]

3:45 Back in the days when cinematographers shot on film, there where camera's where the viewfinder looked through the shutter of the camera. But, one half of the shutter would expose the film, while the other half would be used for the view finder. So, when a still camera took a picture, with the flash turned on, and the camera operator would see the flash through the viewfinder, he was pretty sure that the flash was not exposed to the film.

[@RinoaL]

It's interesting that here recently 30fps videos on youtube such as this one always seem a little jittery compared to 60fps videos. It's kinda like how when I didn't look at a CRT for several years it was a bit hard to go back to looking at one due to the flicker. I think the human eye's "frame rate" varies just like out ability to judge time durations of sounds. You can tell the difference between a 0.15 and 0.20 second sound but you can't tell the difference between a 2.15 and a 2.20 second long sound.

[@FilmmakerIQ]

Watch the follow up video ;) It's not that your eye actually see the frame rate. You see the artifacts of the frame rate.... The same artifacts that can be captured with a camera ;)

[@ScottGard-ed3gy]

Fascinating topic.

[@BobDiaz123]

A while back I did an interesting experiment, I took an RGB (Red, Green, Blue) LED and fired each color one at a time to see at what point I'd see them as all lit at once. At 120Hz (40 times per sec per LED) I saw all the colors, but there was a bit of flicker. At 140Hz, (about 47 times per second per LED) it appeared as if the LEDs were lit all the time. I tried this with both 50% and 75% duty cycles (180 & 270 degree shutters) with the same results. I'm not suggesting that the "frame rate of the eye is 40 to 47 FPS, but assume back in the days of black and white movies, IF someone had a camera that shot 1 frame of Red, 1 frame of Blue, 1 frame of Green, in a repeating pattern, that people would have seen the film in color. I'd hate to think of what the film costs would be to increase the frame rate to 120 - 140 FPS, but it would have worked. Come to think of it, an early CBS System for color TV had a frame rate of 144 FPS.

[@FilmmakerIQ]

Really interesting. Now to complicate that - you're essentially testing the critical flicker fusion of your separate color cones cells independently - smushing them all together to try to get white. So about 47hz per color kinda makes falls in line with established science on flicker. Your idea actually was one of the first color techniques devised by William Friese-Green in early 1900s. Instead of RGB, he used sort of red-yellow - like technicolor 2 strip and doubled the frame rate at the time. The idea of "Sequential Color" was also what CBS's color system proposed in the early 1950s (as you noted). But the modern use of sequential color is in DLP projectors that also have a spinning color wheel. :)

[@Level30Games]

Great video John! Can’t wait to share with some of my film students.

[@ShawnTewes]

Excellent explanation. I was wondering though, rather than using the relative comparison between the eye and the camera to suggest the 10 FPS "frame rate" of the eye, would you agree that it would be more accurate to suggest that the eye has a shutter speed of 1/10? This could be based on how the eye responds to motion blur.

[@FilmmakerIQ]

You can't have a frame rate faster than the shutter speed ;) But remember the mantra.

[@BorjaCampillo]

Incredible. This needs to be in school. Thanks

[@javajack-1]

Oh man, this is an excellent video. I'm am totally in your camp on all of this. In fact I just did a television workshop on the weekend and included portions of what you are talking about. I'm teaching young film students a little about how persistence of vision lead to the minimum 10-12 frames of changing flip book images which ultimately lead to projected film, requiring maybe 16fps to compensate for the flash of black when the shutter closes and the film is pulled down to expose the next frame. My biggest complaint is that young editors don't understand mixing film rates, breaking the constant cadence of an image being displayed. I see this on national ads and I know I could fix it for the editors who botched this, but that issue persists. I'm saving your video for anyone who argues all this stuff with me in the future. So I am a film colorist of more than 25 feature films, thousands of commercials, tv series and have been in the business for over 50 years, so ya, I've got some sense of what's going on. By the way not all films are being shot at 24, remember that Peter Jackson experimented with 48 fps, but it was not well accepted. Of course he's also breaking the 180 degree shutter rule unless he can keep the shutter open for 360 degrees or constantly open at 48 fps to achieve the same motion blur, which is a big part of the film aesthetic.

[@FilmmakerIQ]

Yeah but the Hobbit was 9 years ago ;) Ang Lee is the other... but outside of 5 films in last decade... Everything is shot in 24.

[@javajack-1]

@@FilmmakerIQ yes and isn’t that the great thing about any art form, break the rules and be unique but there is always risk involved.

[@FilmmakerIQ]

It doesn't work ;) 24 isn't an artist choice. It's a medium choice. Going above invokes switch away from cinematic into video. There are artistic applications for doing so but that's more meta than most projects will need.

[@DanDeGaston]

Very well done! I feel like my background in Electrical Engineering and Medical School were both deepend by this amazing movie.

[@MajoraZ]

I appreciate you trying to convey that human vision doesn't fundamentally work in a framerate, and I actually learned a bit from some of the examples... but I feel like this didn't answer what, to me, is what most people are actually asking about when they wonder "What frame-rate does the human eye see?". What I THINK most people mean by that, and what I certainly am wondering, is really better phrased as "At what point is something in our vision for such a short duration of time that the average person won't consciously perceive it?". A practical test of this would be to have displays at varying frame-rates, and asking people if they can tell the difference between different pairs at increasing frame-rates until you hit a point where most people cannot perceive any difference in fluidity between the two displays. I suspect that whatever that threshold is is going to change based on both who is looking as well as what is being shown on the display, but i'm curious if that's something people have tried yet. An alternative approach would be to have increasingly fast objects fly through somebody's field of vision until the person doesn't perceive anything having gone by. Again, though, that will likely vary based on the object's size and other factors (A train moving by at a million miles per hour is still gonna be easily seen if it's long enough that it stays inside your vision for a appreciable % of a second) Of course, this is also really not measuring your eyes or your vision, it's measuring human perception, and at what point enough information is being fed to your brain's visuals systems to where you consciously pick up a blur across your vision or a slightly smoother animation as being a distinct difference or not.

[@FilmmakerIQ]

At what point is something so short that we don't perceive it? That question is already addressed in the video with regards to Bloch's Law. There is no limit to the shortness so long as there is sufficient light to be registered by the eye. You can see a nuclear blast at a trillionth of a second and it would feel as long as a cell phone flash ( though it might be magnitudes brighter). Same concept as a flash and discrete frame rate, as long as there is a flash it will be captured. Next the question if what's the highest rate you can see which behaves as real life.... You cannot have any hope of understanding that question until you understand the fundamentals I laid out here. Concepts like Bloch's Law and flicker fusion are how your can assess that number. I addressed the question of what is the frame rate of the eye. From this understanding you can approach the questions you think are pressing (questions that are not even really about the frame rate of the human eye)... but you cannot begin doing so until you have this grounded in your mind first.

[@noiprocsZ]

brain registers a signal from the receptors on the back of our eyes in about 5ms, depending on the light intensity, you can flash a high intensity image followed by a low light intensity image at low frame rate and you won't see the second image. Eye also have light retention, meaning receptors do not switch off instantly. There is also a fact that triggering one receptor triggers a close one because of the magnitude of stimulation. In other words, we have ghosting just like monitors do.

[@FilmmakerIQ]

Flash two lights in succession is flicker fusion. I wouldn't draw the ghosting comparison. Be very careful drawing inferences. The eye does not actually work like a camera, even a lot of visual processing happens before it's sent to the brain

[@derheadbanger9039]

13:20 so that's why I sometimes notice a flickering light in my peripheral vison - but not when I'm looking at it straight!

[@RobertWGreaves]

Well explained!

[@FilmmakerIQ]

Thanks for that positive note! First one since the moron brigade arrived to roast me!

[@fixitright9709]

Yes the dots on the rotating drum! I don't know if you can but after I watch it enough I can actually make the rotation ,at will, go left or right, it must've been the mushrooms in my younger days???

[@TimothyPleines]

This is fracking fantastic.

[@DethronerX]

Thanks for this video. I wanted to ask, is there any difference in the motion/motion blur, between [watching a movie shot on film in a theater with analogue projector) and [Watching a digital file (on any screen) that's scanned images rendered as video without the physical shutter of the projector? I know cameras already have the physical shutter, but does it look slightly different on scanned images? Can they look more jittery? Reason Im asking is, Im a fan of all mediums, and recently i shot Super8, waiting to send for development and one of them is a reversal that I can project. I'll definitely see the difference if I scanned those images, but Id like to know your experience

[@serraxer]

Where are almost no analog projectors left in cinema. Most of them digital DLP projectors, you see jerks and stutters coz 24 fpc cinema is kinda bad for bigscreens panshots they need dynamics framerates. Just watch Peter Jacksons Two Towers pan shot of Misty mountains at the start of the cinema, it is so jerky and hard to watch on BIG screen.

[@DethronerX]

@@serraxer Thanks! So do you mean, 24fps (or lower) would be more pleasing to watch if projected the old school way? I have 3 projectors, regular 8, 8 and super8 combo and a 16mm one and the Super8s I shot were at 18fps and I dont know what shutter speed the camera is fixed at, its Yashica Elecro 8 LD4, that has auto exposure and no compensation dial and I shot a Kodak Reversal Tri-X, that my developer said would develop it as a Negative, so it would probably be more contrasty and grainy. And I cant wait to see and compare the differences and to decide what looks visually better. I wanted to see a comparison video at normal play and to go frame by frame to inspect if any blur sample images come between normal frames. You probably know what I mean but I cant speak technically. When you add CC Force Motion Blur, you can change blur samples that appear between frames, that look like two overlapping images with slightly varied action and I think those make it pleasing to look at in normal speed for lower frame rate. Sorry for long message, knowing technical words would have shortened it : D

[@DethronerX]

@@serraxer And yeah you're right about panning shots and I think the analogue projector may fix that with motion blur or digitally, they'll have to add it. 24fps is still there in modern cinema, its just 23.976 or 24, depending on the camera with 180 degree shutter, unless otherwise for an effect of a scene. Faster framerates havent worked, but I dont know about framerates on very recent movies shot on digital, but I do know that the Will Smith movie and Hobbit didn't do so well because they start to look like a soap and I think its been talked on this channel about those movies and how speed, like chases, dont look fast enough. For panning shots, they could change framerates though, like theyd decrease for some fighting sequences

[@FilmmakerIQ]

My next video will be on flicker. Yes there is a big difference between a computer monitor and a projection with actual shutter. 24 fps is not bad for the big screen (how could it be if we've gone all these years). Flicker (and low light levels) is why 24 fps looks smooth in a theater.

[@DethronerX]

@@FilmmakerIQ I saw your recent Live session and you talked about Pi, with the contrasty look. I read somewhere about Tri-X and Bolex and now just checked IMDB: Cameras: Aaton XTR Prod, Canon and Angenieux Lenses, Bolex H16 (Ultra 16 wide and blown up to 35mm) and the film:(Eastman Plus-X 7276 and Tri-X 7278)

[@gluttonousmaximus9048]

Please to a vid on MGM and WB's studio history!

[@jfeeney100]

BRAVO! Love the video. Something I would like to see studied with Visual Perception. Back in 2011 I got involved with a study in 3D cameras. I had a setup that was absolutely amazing, and shot some material. We developed ways of handling L / R pairs of videos, and combine them into a single 3D L/R video. Sometimes when I was working with the equipment, I would accidently reverse the Left and Right eye as displayed on the monitor. What you saw when you looked at the monitor was very subtitle, but the effect would grow the longer you stared. You would initially notice nothing a normal 2D picture (not 3D). Slowly little areas of the video would start to take on a 3D ness, but it looked inside out. After 10 - 15 seconds of watching the reversed video, larger, and larger areas started to take on the inside out 3D effect. It was very interesting. I then stopped the video, and looked at another part of the room (reality), and for about 10-15 seconds reality had this inside out 3D look. It was a very strange experience to view the video in this way. I wonder if anybody else has done anything with reversed Left Eye / Right Eye perception?

[@FilmmakerIQ]

That sounds absolutely trippy!

[@GiantBean]

This is fascinating and while I have long thought the human eye has no frame rate as it has no shutter and works on a biological system that interprets, predicts and even fabricates inputs to create vision in a very different way then machines, I would now love to see a video on VR headsets and why motion below about 74 frames often make people sick.

[@FilmmakerIQ]

I think it has to do with inherent lag of head movement and visual perception. What I've read it's the same thing as motion sickness from sitting in cars.

[@zetacrucis681]

You can see the 3:2 pulldown judder in the NTCS DVD version of the famous train station crane scene (Jill's arrival) in 'Once Upon a Time in the West' as the camera rises over the station house to reveal the town. Same scene in the PAL version has no judder but the pitch of the soundtrack is too high (coz 25fps). Before blueray was a thing I was annoyed enough with both these artifacts (for that particular film) to buy both NTSC and PAL versions of the DVD and remux the NTSC audio with the PAL video slowed down to 24 fps.

[@FilmmakerIQ]

No you can't. I never went through with the 3:2 pulldown video because my experiment demonstrated that you don't see frame by frame. (I tried a bunch of different ways to try to tease out the effect) Even with 24 frames your eye takes in about 3 frames at a time, and when I discovered Bloch's Law, there's no way you're discerning the pulldown. So whatever was causing the judder of that one shot, it wasn't 3-2 pulldown.

[@zetacrucis681]

@@FilmmakerIQ You can see that it's limping compared to regular 24 or 25 fps. You can also see when motion judders with mismatched frame rates, e.g. 24 fps footage is shown on 50 Hz refresh display. Does not mean you're seeing 50 or even 24 individual frames but you can still see that movements like pans and tilts aren't smooth but have a displeasing stick-slip quality to them.

[@FilmmakerIQ]

Again, it's not 3:2 pull down. It could be an encoding error caused by mishandling the "dejudder" which would result in an accidental 12 frames a second. That looks like it's struggling. I recently had to deal with that in a video someone sent me. But the reason why it's not 3:2 is because you think straight 24 looks good on your 60hz display. Also as I demonstrated in this video, you don't see individual frames but collections of frames. 24 on a 25 hz display is different because the pull down for that is a hiccup even second, that's much more visible.

[@zetacrucis681]

@@FilmmakerIQ Okay, thanks for the clarification. Sounds like an instructive experiment to do for myself: convert 24fps to proper 3:2-pulled-down 60 fps and see how it fares.

[@zetacrucis681]

@@FilmmakerIQ Did the experiment with the 23.97 fps bluray version of the same scene and you're absolutely correct: can't see judder in the 59.96 fps 3:2 pulldown conversion. In fact, synced up side by side on a 119.92 fps refresh display sometimes I thought the original was smoother, sometimes the converted version was smoother, so basically I can't tell.

[@thewonderfultartiflette4733]

6:00 but cursor phantom array isn't the only way that we can differentiate between 60 and 144. Just play a game at 60 and 144 and you'll see the difference, but in this case, the camera will not be able to catch it, how do you explain that ?

[@FilmmakerIQ]

Sure the camera will catch it. Same thing as the phantom array effect. Here's an article with the camera catching it: blurbusters.com/the-stroboscopic-effect-of-finite-framerate-displays/

[@jakubmicuda]

Great video.

[@MarkoTiosavljevic]

Doing measurements of refresh rate on LCD monitors vs Human Eye vs Camera shutter speed. It's like going to the moon to measure your weight :)

[@FilmmakerIQ]

I wish I weighed here what I would weigh on the moon

[@MarkoTiosavljevic]

​@@FilmmakerIQ hahahah :) anyway ... rly a good story, Theory is totaly ok... i am just concerned about numbers ...

[@jccarlyle]

One saying I really appreciate is "all models are wrong; some are useful." The concept of frame rate is also a model; a useful one, sure, though I still believe that insisting that there's no inherent frame rate to human vision is probably for the best, even if we can perceive some difference from, for instance, playing games at a high frame rate. Even the concept of frame rate breaks when we consider that no display is capable of presenting all lines or pixels at exactly the same time, but that doesn't mean we will perceive flickering. Great video!

[@FilmmakerIQ]

The concept of frame rate in human vision is NOT useful. It's the opposite of useful.

[@jccarlyle]

@@FilmmakerIQ It *is* a useful concept when you're talking about display technology, and that is obviously related to human vision. I suppose you're right, though, discussing human vision strictly or mostly in these terms is indeed harmful.

[@FilmmakerIQ]

No it's not. LoL In fact it's absolutely detrimental to understand display technology! It keeps you solidly boxed into a limited understanding... For example, let's talk about BFI. How can you insert a black frame flicker of say 120hz if you believe your eye can see 240 frames a second? Let go of childish beliefs and join the dark side ;)

[@AdrianBacon]

Interesting video. I agree that the human eye does not have a frame rate like what we would describe with a video camera, but, I do tend to think that like what you covered, there are some interesting numbers that could be conflated with a frame rate if somebody wanted to argue that. My view on it is that there are some frame rates on display devices that perceptually are more pleasing to the human eye. Sports on TV at 60 fps is one example. It just looks better at 60fps, and much like gaming, would probably look better with more fps. Gaming... more is better. I don't tend to think that the eye has a FPS per se, but rather has a rate at which it can detect movement smoothly, the lower end being down in the 10-15 FPS range, and the upper range a lot higher than that... How much higher? I think that depends, but to your point, the eye is not a video camera, and while it has a lot of similarities to a video camera, it does not work like one, and therefore does not have a "FPS".

[@williambell4591]

PROPS for this, I enjoyed this video immensely!

[@random_gamer_09]

Mind blowing video.

[@FreyaFromAsgard]

Are you maybe planning to make a video about connection between frame rates of the VR googles and human eye? I actualy looked up that video trying to understand why people are feeling sick while VR goggles have lower fps. Can you maybe explain that topic?

[@FilmmakerIQ]

I'm planning on doing one talking about the frame rate needed to completely get rid of any artifacts of discrete frame rate (simulate actual reality)

[@FreyaFromAsgard]

@@FilmmakerIQ Can't wait for the video then! I'm writing my bachelor's thesis about making most immersive VR space taking into account possibilities of current technologies. That would be very helpful! Thank you for that video and I'm waiting for the next ones too! 😀

[@rowxah77]

I am a gamer i play on 240 hz monitor... i get what your saying with 10 fps but please explain to me this, on reaction time tests i can get a consistent 135ms click so by the assumption i see at 10hz does that mean my muscle can react in 35ms because that sounds outrageous, and i know i see the box go geeen at about 20ms

[@FilmmakerIQ]

Not all the cells in eye are synced up to 10hz even if the individual cell has a reaction time of 100ms. Watch the follow up: kzbin.info/www/bejne/bavOqmCporeSkLssi=9SQo6Stn9Dmn5wk3

[@Lalovoe]

13:33 so THAT'S WHY I notice a slight bit of flickering on my Nintendo switch OLED in the dark when it's in my peripheral vision

[@FilmmakerIQ]

It's why stressed out people in office buildings get even more stressed out by the overhead fluorescents lights! ;)

[@thedroporg]

Thanks!

[@DaddyBeanDaddyBean]

On old-school CRT monitors, I can *absolutely* see the refresh rate of a static image at 60hz - it flickers, faint but unmistakable, and can trigger nausea and ice-pick headaches in a matter of minutes. Bump that refresh rate up to 72hz and I can no longer see it, and can sit on front of it all day long with no I'll effects. Note LED/LCD monitors are different and don't actually HAVE a fresh rate of a static image - all the pixels that are lit, just stay lit, until the image changes. The flicker on a CRT is due to the fundamental nature of how an image is drawn on a CRT screen, with each pixel being lit up as the beam passes over it, and then immediately starting to fade/decay until the beam passes over it again 1/60th or 1/72nd of a second later.

[@FilmmakerIQ]

Yup, but that's flicker sensitivity, a little side step from what I'm talking about at that point.

[@DaddyBeanDaddyBean]

​@@FilmmakerIQ fair enough. 😎

[@thehonestdude1067]

Very interesting 😄

[@j7ndominica051]

When I drag a mouse pointer, I do see multiple pointers. Part of it is caused by low response time of the LCD. Rather than a mouse pointer, they could grab a window with something to read and see how well they can continue reading it while in motion. Scrolling text on news stations is impossible to read at internet framerates. In games we usually pan the entire screen past us very quickly all the time when looking around. This rarely happens on video. And we are also an active participant in the scene where the time delay between the input and the result matters. When looking at a rotating wheel in a passing car, there are moments when the wheel appears clear. I think it is because the eye catches up the position of the wheel at a similar speed to its rotation.

[@FilmmakerIQ]

No, you see multiple cursors because the frame rate is not high enough to create motion blur. You could have instant response time and still see multiple cursors... You will continue to see it until you get into the realm of 10,000 or more fps. The moments where the car wheels appear clear may be due to saccades, essentially you eye stops gathering information as you eyes move so as not to create a sense of motion blur as you eyes dart about.

[@DescendDab]

To me, the only upside of high refresh rate monitors is when something on-screen moves very rapidly. You would see the path that the rapidly moving object takes, more clearly.

[@41divad]

Excellent

[@WrathChild-NZ]

good explanations, Now a video on what the F stop range of the human eye is lol

[@FilmmakerIQ]

That's actually pretty straight forward to calculate. I've seen it to be between f/2.8 and f/8

[@WrathChild-NZ]

Oh nice!

[@jbstepchild]

During the candle shots I had a longer trace around the flame in the 1s shot vs the 2s an my eyes preferenced the 2s so the 1s seemed darker

[@luckyumbasa417]

Very interesting. Honestly makes me wonder about some of my fundamental assumptions about why even stable low framerates look, not just feel, so awful in games but great in hand drawn animated film. The only game I'm aware of that even tries a low framerate approach to animating is from the GDC talk about Guilty Gear Xrd's Art Style. While it works amazingly there I havent been impressed with the few 3D animes I've personally seen try this...

[@denizkendirci]

19:50 it doesn't look like going one way or another and switching directions time to time though. instead, it does seem to be going both ways at the same time by splitting at the middle into two opposite streams and directions of motion. or maybe it's just me, i don't know.

[@FilmmakerIQ]

The mirror takes a little more time then I gave you in the video. Go to the Wired video to see more of it.

[@denizkendirci]

@@FilmmakerIQ Will do, thank you.

[@Benjamin-nf2ir]

Amazing presentation. Better than a college course. You are forgiven the click baity title. 😊

[@TiagoTiagoT]

For gaming, there's more than just aesthetics involved, as it is an interactive medium and response time and detection of details can make a significant difference. It is more understandable things may get more subjective when it comes to non-interactive media though.

[@FilmmakerIQ]

That's still an aesthetic ;)

[@BrianBaastrup]

If the controller latency is low enough. Then framerate becomes redundant. Unless you have health issues relating to it.

[@TiagoTiagoT]

@@BrianBaastrup Among other things, it's about reaction time.

[@ArchAngelAzmuel]

From looking into it and testing things myself i stopped saying brain or eye fps. While they eye can see nearly continuously there is a preception limitation in the brain. That is there is a point where you wont really be able to process which fps is faster, that point is around 300ish fps from experience. That said while solid motion may only need 24 fps, when i went to see DND recently i noticed the image stuttering, and while i could ignore it it was occasionally distracting. I use a 240hz nonitor for fps gaming and most monitors today are set to 60hz shouldn't for ease and comfort of the viewer at least shift closer to that. All that said i might be a special case as im photosensitive and often notice if a flourecent (the few that i can find) is flickering while to others it is solid.

[@FilmmakerIQ]

Yes you can process screens above 300fps... Wiggle your mouse harder. To get it completely gone you might need to get up to 10,000 fps and beyond! ;)

[@NicholasIstre]

The main reason pro gamers seek high frame rate monitors is to shorten the time between input and changes on the screen. This is definitely perceptible between a 90 Hz screen and a 144 Hz screen (and likely higher with diminishing returns, see the Linus Tech Tips videos on this). Without that interactivity? I probably wouldn't be able to tell the difference between 90 fps and 144 fps in a blind test. And I'm perfectly able to watch media at 24, 30, or 60 fps just fine, even when it's right next to 144 fps gaming (Think pre-rendered video like in some parts of Kena at 30 fps) without even noticing unless I look for it. But I will immediately notice once given control and move the mouse at those lower frame rates. The other side of pro gamers shortening the OODA loop is seeking low latency keyboards and mice. That has no bearing on cinema or tv shows. Anyways, great video!

[@FilmmakerIQ]

Or is it placebo? The monitor feels snappier, so the gamer's play is snappier. Lots of folks bring up that the gains might be overshadowed by the limit of human response time. Still you can see improvement in gameplay even if that's not a direct result of the actual frame rate gain. Given everything I've read so far about visual perception, I think the latter is true. Doesnt mean the high fps monitor isn't serving a purpose, but that the part getting there is different than you think.

[@xan42O]

Loved the vid, especially as a medical professional, filmmaker, and gamer! The flicker fusion threshold (60-90 Hz) seems to be the best "answer" for the question posed as that is literally the basic principal of motion in cinema. One thing I would add for gaming is that higher is always better due to input lag and how terrible it feels to play with lag.

[@FilmmakerIQ]

The flicker fusion is going to play a big role when I talk about how flicker actually improves clarity and enhances motion in lower frame rates. We've been accidentally using flicker for 100 years with projectors and Cathode Ray Televisions. Now that technology has moved on, we need to reintroduce the flicker!

[@xan42O]

@@FilmmakerIQ i cant wait to learn more about it!

[@akillersquirrel5880]

Yeah, flicker fusion threshold plus phantom array are IMO the strongest points for the perceived experience of high frame rate monitors. Subjectively, I know that my flicker fusion threshold is at least partially above 60Hz - cheap LED Christmas lights (which strobe at 60Hz or whatever your local power grid's frequency is) look steady in the center of my vision but flicker outside of the center.

[@FilmmakerIQ]

Wave your hand in front of your face when sitting under a fluorescent light that flickers... Phantom Array!

[@GangsterFrankensteinComputer]

Does this also explain why frame rates lower than 90 can cause VR sickness?

[@lindigj]

We don't see in framerate, as noted at the start. The way we do see is more akin to a 10 hz monitor that doesn't have any way to sync frames. Without any way to sync, things like flicker and judder will be detectable with high motion if the eye isn't being bombarded with an excess amount of frames to ensure there is never any overlap during the high motion sequence.

[@08wolfeyes]

Would the speed of light perhaps be considered the limit of our frame rate, if we want to call it that for human vision? The reason we see anything is really down to light, which has a speed limit.

[@FilmmakerIQ]

Not speed of light, but Planck time could be considered the "frame rate" of reality. But that's so much higher than anything we can physically experience.

[@sauzeefy]

good video, but your monitor demo there around five mins in doesn't properly illustrate anything for either side.. the pixels on an LCD can't change instantaneously. It's true that the two monitors refresh at different rates, but the pixel response was likely similar on both at around ~50ms. Naturally, this would mean that those after-images would appear no matter the framerate. This also would explain the three dots you saw in your demo shown shortly after.

[@FilmmakerIQ]

No. This has nothing to do with afterimages. You aren't the first (or last to make that mistake). Any exposure longer than the refresh rate will result in multiple exposures regardless of the G2G. Exposure is ADDITIVE. The refreshes on the monitor ADD together while the sensor is exposed to light. Also the camera here is static, so it would not be affected by G2G. You need a pursuit camera following a "moving object" on screen to see what you're talking about. That's what that camera on the slider in RTING videos are.

[@kevinfeathers7432]

Wait until you tell them all what happens to colour when you turn the lights down. Oh boy they are in for shock and will learn just how much the brain "paints" the world.

[@crazydud2432]

The spinning drums wouldn't film randomly for me even when staring at the dot, but if I imagined they were spinning in a specific direction they suddenly were. I could get the spinning drums to flip simultaneously or desync at will. I never thought the eye had a definitive frame rate, the idea always seemed rather silly to me. It takes a certain amount of exposure/time to form a fully complete picture maybe that is around 10 hz. The eye can still see things much faster than that, at the cost of some of its detail.

[@mediamass1404]

21:23 if it does happen but rarely its probably something to do with the minds focus independent of the vissuals, if you interpret the mind as a holon of interactions between difurent systems that makes the most amount of sense here

[@mediamass1404]

11:08 it still happens but its weaker, maybe its youtube but I still see the right as slightly brighter, maybe test it with variouse difurent speeds

[@FilmmakerIQ]

You also may have a slightly longer sensitivity. These numbers in psychophysics are number hard lounge numbers ;)

[@matheus5230]

The illusion you showed around 19:40 of the video works for me, but only if I actually make the effort to see it, like looking more to the side. I feel it would be more effective in real life. Cool nevertheless!

[@serraxer]

Well this video was made at 24 fps so in real life you can percieve much more. Just look at some long moving billboard of moving bus in movie vs real life you will catch difference. Human eyes can adapt to HFR, more you watch better you can catch and yeah we have some boundaries of percieving 240 - 500hz I think.

[@nimbulan2020]

A pretty interesting analysis here. Unfortunately I ended up here from a clickbait article that used the "10 fps" number to try to claim that higher framerates and cutting edge tech are entirely useless.

[@FilmmakerIQ]

Is that why everyone is so angry with me? I never said higher frame rates are useless. The 24 fps camera can obviously see how higher frame rates would be beneficial

[@FilmmakerIQ]

I think I found that article but it wasn't claiming 10 FPS was the max. (Digital Camera World?) It instead posed a question of 10fps was wasting money... And then if you read it, it actually explains pretty much verbatim what I explained in the video. Remember the rule: if a title of an article contains a question, the answer is usually no

[@nimbulan2020]

@@FilmmakerIQ Yeah they did cover the points in the video, it was just presented in a way that really seemed to emphasize the initial 10 fps and better tech is useless premise, at least to me.

[@FilmmakerIQ]

That title is more clickbait than I'm comfortable with. But I think there's still a lot easier to understand certain aspects of human vision if you take at least a very loose conceptual model that human vision is around 10hz. Things like motion blur, connection with where apparent motion begins, where the subjective wagon wheel effect start... it's all related. Just don't draw extreme conclusions.

[@yiravarga]

Now I see why 24 fps looks good for video. The exposure time per frame, is longer, basically a static image held for longer, improving perceived brightness and clarity for each individual frame. Like the mouse cursor, higher fps equals more cursors captured for each exposure time. Less cursors captured per exposure, means each individual cursor would be sharper. Each cursor is far away enough from each other that there is no difference in how each individual cursor looks, but move them slower, and the higher fps one, the cursors overlap, creating a blur, which then gets captured as a single frame by the eye and brain. I’d love to see what a 24 fps video shot with each individual frame at 1/1000 or more shutter speed. What would video look like if each frame has little or no motion blur? What if each frame was a super sharp clear image, as clear as a photo?

[@FilmmakerIQ]

You'd get much more stutter. One mistake you're making is that the cursors are just as sharp in the 60 hz capture as they are in the 144hz. The frame rate doesn't affect the sharpness, it affects how many cursors there are. There will be a point of you get the frame rate high enough that there will be so many cursors they eventually blur together but that's like 10,000fps or more. With the conceptual approach that the eye sees 10fps, a monitor showing 24 fps would give you an experience of seeing 2.4 cursors. So a no-motion-blur image would stutter pretty hard because you're summing up 2.4 frames at a time. Think the opening of Saving Private Ryan

[@yiravarga]

@@FilmmakerIQ thanks for this informative reply! At the numbers we’re working with, the demonstration of differences with the mouse cursors aren’t apparent. Also, I’m sure neurons have a time length before they send another action potential. Each eye receptor is on its own clock cycle, so there will still be motion blur, but no frame rate, just analog smooth motion. I’m glad you mentioned that in the video. I just tried this with my lg oled. It has a mode for black frame insertion. I ran a video game, because it’s like the mouse cursor, just a clear sharp image per frame. Wow, it’s very unwatchable. Even with black frame insertion off, it’s still very jarring and uncomfortable (24-30 fps). So… iPhones, and other devices probably do some kind of motion ailising? This really shows how much more complex modern displays are, and the extra tricks needed for oled. This is the most complete video of how we see motion I’ve seen so far, great job!

[@FilmmakerIQ]

Yes extra tricks are definitely needed. I used to think BFI was the solution for 24 fps but I have to revise after extensive testing. It's actually what happens between frames that matters. Don't think in terms of neurons with clock cycles. Neurons do not have clocks, that's why we don't actually have a frame rate.

[@gruntaxeman3740]

I think that choosing frame rate correctly is very important on every video production, and it is somewhat related to human visual system. For "cinema", or lets say "visual fictional story telling", 24fps is good "normal" framerate for these reasons: -It hides acting -It hides "animation" on VFX -Higher fidelity frames because video is usually compressed -Allows effects for storytelling by shooting lower framerate or increasing shutter speed. -6fps/12fps/24fps/30fps/60fps/120fps.... what are these numbers? Nice fit for sexagesimal system. Babylonias used that while it divides nicely. Framerates today seems to be "natural" standard in video productions to get numbers that are easy to dive. -Flawless audiosync with 48kHz sound So I don't think that 24fps is disappearing. But when shooting documentary, something that is "reality" and not "acted", higer framerate is preferred. Videos in web like KZbin benefits from 30fps or higer. Also video blogging benefits while camera is in hand, there is easily high panning speed that looks ugly on 24fps. Broadcasting documentary/news 50fps or 60fps is often optimal but in streaming KZbin video compression degrades frames much, so 30fps is usually better and 24fps for visual story telling. Game videos of course works same framerate as games. These actually works games too. Motion blurred 24fps actually can work in games to hide "acting" in motion capture or hide inaccurate physics in animation. Except when there is action that requires reaction time, camera is moving and panning so it should be higher like 60fps on action games. Slow paced, slow panning camera can work 30fps. We see rarely succesful mixing of framerates because it causes contrast between shots so changing framerate should have purpose, like visualize LSD trip. Some games may possible benefit to enhance contrast between cutscene and gameplay by changing camera angle, and framerate.

[@FilmmakerIQ]

I pretty much agree with everything here but I think 24 has a place in documentary. You see almost every HBO documentary shooting in 24 and I think that's because it gives it a higher dollar cinematic look. 30 FPS has more of a live in the moment feel. Even though broadcast is 60 you'll find that most broadcast productions are actually still 30 ( or 25 in PAL)

[@gruntaxeman3740]

@@FilmmakerIQ I actually have not seen any HBO documentaries, recommend something? :) But sure, there is use to 24fps in documentaries. Like some medieval history documentary where historical scenes are acted so 24fps is best here because actors don't really hit others on swords or morning star, that is illusion and story telling :) Documentary may be interesting for framerates if there is mix of acted story telling and having television host climbing on some mountain or jungle in real location. That can make perfectly sense to mix frame rates of 30fps reality and 24fps story telling, and even add more contrast by having some color grade on acted visuals and then come back to shooting televison host with neutral colors. I know that broadcasting is shot on 30fps and 25fps but in my understanding they mostly move back from 60fps and 50fps, because higher framerates were shot in interlaced format (analog videocompression) and now same content is also streamed over Internet where progressive 30fps and 25fps works better, so they likely want to avoid conversion hazzle and keep similar look in all formats. That would be interesting if someone figures out new ways to tell story (other than documentary) without acting. Actors, or more accurately "models" unaware of shooting. Candid camera did this or purpose and Truman show made idea to satire. If someone figures out new ways to do that, that would work high framerate. Interesting approach to accomplish this would run events on simulation to record all reactions and emotions. from models. Looking someone playing videogame does not look bad on 60fps because it is not acted, gamers reactions are genuine. But if someone build simulator to capture real emotions and reactions, and make movie by placing virtual cameras on interesting angles and cut material to movie, that can be interesting and actually work on high framerate. This is possible or almost possible in principle with huge development resources. This kind of stuff has been done on some Star Wars movie to make projected background to update realistically actor moves contollers on spaceship cockpit. Or director shooting gun to capture real reactions, or actors actually drunk shitfaced to play drunk. Still in 24fps because movie is still acted, planned staging and blocking. I would like to see that someone manages to make whole movie without acting, only real reactions and emotions captured, real stunts, all reality or simulation. That can work high fps.

[@gruntaxeman3740]

@@FilmmakerIQ Oh, I forgot.. Something like 15year ago I shoot footage from airsoft combat, running in bodyarmor and shooting Sony VX-2100. That was actually beautiful high fps combat footage while models where not acting much while avoiding projectiles, and cameraman too. If this is done with accurate gun replicas, mix sound effects, add VFX muzzle flashes but real smoke grenades and more cameras, head mount GoPros and so on... That may work beautifully as a tool making high fps, immersive story telling.

[@FilmmakerIQ]

some good HBO docs: The Inventor: Out for Blood in Silicon Valley (about Elizabeth Holmes scandal) Going Clear: Scientology, Hollywood and the Prison of Belief Netflix has quite a few as well, although you can make a case that they don't have to be 24, aesthetically they just have a more serious look in 24. Shooting 30 makes it feel like a TV newsmagazine.

[@gruntaxeman3740]

@@FilmmakerIQ Thanks, I just watched Elizabet Holmes document. Large part of document was showing old film material and also artistic shots so 24fps makes perfectly sense. I don't want to see framerate to jump without reason so it is better to pick one that works. I think that 24fps is best for acted performances, animation and it is artistic tool. And that covers almost 100% of cinema. But in this documentary, those handheld interview shots where interviewer and cinematographer walking without ninja walk... Those shots easily sucks on 24fps while there can be accidentally too fast panning. In my opinion, fast drone flights, benjijumping with GoPro, rotating crane shots, walking & handheld interview shots without planned staging and blocking, those crazy CGI shots where camera is spinning and flying... Those actually look much better in 30fps while acted performances may start to look like school play. So shooting whole movie/document 24fps is then obvious choice. I have not seen yet good examples of changing framerates during movie that makes sense. In fact I feel changing of camera between shots often as annoying, visible distraction.

[@1dgram]

Could you theoretically have overlapping exposures, hence a shutterspeed longer than the frame rate? I bet it would look really dreamy!

[@1dgram]

(Really it's one over the framerate to get what's normally the longest possible shutter speed, but we all know what you meant.)

[@FilmmakerIQ]

Not possible when recording in real time. You could only do it in post. I still have an idea to make a video talking about what it would look like if you summed up a few frames. It would actually look like an echo - multiple frames overlapping each other.

[@1dgram]

@@FilmmakerIQ I was thinking animated graphics, but if you can retain a smooth motion blur in each frame then it should be possible to add this effect in post from something shot at a 360 degree shutter angle.

[@FilmmakerIQ]

Yes, 360 would be how I would do it. As I said it would be an echo. Because the motion blur would end after the next frame began. Essentially each frame would back up in time... The effect in After Effects is actually called echo

[@SchardtCinematic]

I even have trouble seeing much diffrence between 1080P and 4K. I have a 4K TV as my monitor and 1080P and 4K look about the same to me.

[@FilmmakerIQ]

It's pretty common. For a lot of viewing experience somewhere between HD and UHD is where you really stop seeing the difference.

[@SchardtCinematic]

@Filmmaker IQ Yes I definitely agree. I used to be amazed by 4K when first seeing the demos in at Best Buy years ago. Until I realized it was more about those ultra sharp lenses they were using. Not the 4k. Because I looked up some demo video used in diffrent stores on you tube years ago and even playing in 720p on my old Windows 7 laptop the video still looked amazing. I use a Canon R7. But most of my video I still shoot at 1080p. If I'm nature or on a trip then I'll switch to 4k Fine to bring out more background detail.

[@davidhasbrouck7948]

Sometime I would like to watch a movie or just a video at a high frame rate like 240 fps to see how pans and motion look. Our eyes don't really pan and if we track motion there is no blur so I would like see what it is like to watch something that let's my eyes do the blurring not the camera. I wonder if there is an uncanny valley of nausia.

[@matheus5230]

If we go by Blur Busters, then 240FPS is not enough. We would need to go to 1000FPS at least, tens of thousands of FPS actually, to create motion that would be absolutely indistinguishable to real life.

[@davidhasbrouck7948]

@@matheus5230 wow I didn't realize it was that high. I wonder what the frame rate would need to be able to track moving objects though. I take photos of my kids running around at 1/250 and it stops the motion well. But maybe that doesn't translate to video.

[@FilmmakerIQ]

Try shooting 60fps without any image stabilization - that's already in the realm of nausea :P

[@moonasha]

I think you can just boil this argument down to, the human eye doesn't have a shutter.

[@FilmmakerIQ]

Well yes, but neither did the old video cameras and they still had a frame rate. Also get your troops away from my border Miriam