What is the "frame rate" of a human eye?

[ReiAyanami]

the eye sees infinite FPS, just the brain cant process it all. thats why flys cant react to fly swatters

 

[merlocka]

Originally posted by: wfbberzerker

Originally posted by: Mingon

yes...it is 24 images per second, not 60. i dont know if that is the same thing as all thi FPS babble, but i know for sure we cant see 150fps haha that is hilarious

Have you even tried either of the 2 programs posted ?

i wouldnt trust that program...

Don't trust it because they might be virus's or because they obvioulsy illustrate the difference between 30fps and 60fps (or higher) when rendered on a PC?

Everyone in this thread should play with those two applications for a while. The results may impress you.

My personal opinion is that there are too many factors in the interaction between the eye and the brain to place some general max perceptable "fps" limit, and even if you could it would be different for each person.

 

[dxg]

60Hz, Its hurts my eyes. I cant look at it for long, its painfull to look at.

75Hz, Its ok. But flicker is still quite visible.

85Hz, Smooth. Flicker is pretty much invisible.

120Hz, Very solid image. Looks extreemely good.

I can easily tell the difference between each one.

24 is the lowest possible rate at which things look smooth to humans, but the more the better the image will look.

A game running at 100Fps looks much better than one running at 24Fps.

The human eye is analog.

 

[Somecallmetim]

Something that comes to my mind is the fact that al the infromation given so far is using the eye as the stationary focus. As we track movement in real life, the eye moves with a angular velocity that helps to compensate for any "motion blur". As the eye moves, the the mind actually sort of subtracts the speed of the object and the speed of the eye to slow down the object so our brain can understand what the object is. You can see this happen when you turn away from your screen rapidly, so the brain purposely "skips" frames so that it can resolve the image. Now if the eye is stationary, as is the case (im assuming), the picture blurs, but the movment of the eye stays the same, so the brain has the "object" on the screen at a acertained velocity, with no velocity for the eye, so baseline the eye can resolve just about any fps. As far as any actual values, I find it hard to put a certain value on the eye, as it seems that the brain is what actually slows or speeds up recognition of an object through the eye, but it must be that magical frequency at which the brain runs I assume.

 

[dmcowen674]

Good post Tim. Bottom line is we can't measure the human eye in purely FPS. We suffer from the same fate as basically amounts to the old phosphorence type monitors, that we have burn in effect.

 

[Trake2011]

So the thing that people aren't factoring into this discussion at all is that there is a difference between what you see and what your brain interprets. For example, the optical nerve presents a blind spot in your vision that your mind otherwise glosses completely over (e.g., you don't see a blind spot). It's a spot in your retina where rods and cones are gone. No one sees this because the brain interprets data. Another example is that the physics of the lenses in your eyes present a physically inverted shape to your retina. Your brain interprets this so the world makes sense. People attribute all of this to the eye, but the eye doesn't process, it just triggers a cascade of polarity down an axon to the centers in the brain that do process.

It takes about 5 to maybe 12 milliseconds for an the electrical impulse in a neuron to fire and reset depending on a lot of what-ifs. That means an individual retinal nerve, which takes a response from ~100 to ~1000 rods/cones depending on where in the retina you are, can fire every ~7ms on average or about 140 Hz. There are about 1.3 million retinal neurons available to send this signal. The don't fire in concert to present one frame per cycle.

Great, so what? Well, you can think of the eye as having a maximum resolution and each retinal cell as sending a continuous 140 Hz stream of information and each pixel refreshes independently and with no particular synchrony with the pixel near it. The brain then interprets this data to make sense of it. There's no frame rate to speak of because our physiology doesn't work that way. Our brain is interpreting a continuous stream of information. We can see a picture lasting 1/220th of second because it provided enough light to cause enough neurons to fire that our brain could analyze the information. A 24 fps video provides just enough information that our brains can fill in the gaps to makes sense of it and interpret continuous motion. 220 pictures in a second would provide our brain with too much information to articulate and thus we would see a whirl of images, or maybe just blurred light. Point is, you can't exactly consider your eye as having a frame rate the same as your brain doesn't really have clock cycle like a CPU.

I found my way here because of new 240 Hz TVs. The fact to all of this is that if it looks better, then you're eye can resolve the difference. You might not know why, but perception is reality and if you perceive it, then it's there. Right...?

 

[wuliheron]

The eye itself doesn't process frames. The brain works like an FM radio with each of the senses (except smell) sending analog signals to the brain where they are then compared against built in oscillators. The differential between the two provides much faster and more accurate information then sending frames back and forth. For vision the oscillator signal moves through the brain from front to back, for hearing another one goes side to side, and for touch it is top to bottom.

There are all kinds of optical illusions that can fool us and most of the data the eye produces never even reaches the brain, but we can still detect even a single photon or the slightest motion thanks in part because it is the differential being measured before the brain itself ever plays with the data. You can get an idea of just how slow the frames your brain produces are by watching cars on the highway out of your peripheral vision. That stutter rate is roughly the frames per second your brain produces and as much as 80% of what you see is just your brain filling in the gaps.

Others have already implied it here, but anywhere from 120 to 240 fps is ideal for a monitor. The only reason most monitors don't produce that yet is because of price and our vision is so good and our brain can fill in so much we can compensate for significantly less then ideal fps without it ruining the experience.

 

[Jeff7181]

The human eye (and human body, for that matter) is not a digital system. There is no hardwired limit. Differences between people (biology, genetics, training) also come into play.

While it's safe to assume any human eye can't tell the difference between 1 and 2 million frames per second, when you get down around 100 it's a grey area. Military personnel are trained to identify targets quickly. One of the ways they used to do this is to use a projector to flash a image, or silhouette of the target for a fraction of a second. My Speed Reading instructor used the same method to get us to recognize the "shape" of words rather than to focus on the word and its individual letters. I got to the point where I could identify a string of 3 words displayed for 1/500th of a second fairly consistently.

 

[komatta]

Bump after 9 years? Interesting...

 

[A5]

Bump after 9 years? Interesting...

Yeah. A really good 1st post, though

 

[JEDIYoda]

i think i remember reading that the human eye sends about 24 images/second to the brain (it was either that, or 60, cant remember). im pretty sure that in addition to correcting the perspective (images are sent to the brain upside down, since the lens of your eye flips the outside world), it also blurs images if they are moving too fast (somewhat like the motion blur on TV).

No..it is actually 32 fps....

 

[SunTzu1983]

http://www.100fps.com/how_many_frames_can_humans_see.htm

 

[Red Squirrel]

The eye is like a high speed camera hooked up to a DVR that can only process a limited amount of data per time unit. The camera itself may be able to capture very fast but the DVR can only record so much of it.

For the leds experiment, reminds me of projectors. When there is an image projected if I move my head fast enough the projected image actually turns into a blur of color.

I think an influence as well is how much activity is happening in the image. When watching a movie that is probably at 24fps you don't notice the frames because there is lot of movement going on so it keeps your brain busy processing that, but if you are staring at a solid color wall with something flashing at 60hz you'll notice it. Take those cheap LED Christmas lights for example. Pretty sure those flash at either 60hz or 120hz depending on if they use a bridge rectifier or not. Most don't convert to a smooth DC curve so it's pulsed DC being sent to the LEDs. Reason you don't see this with incad bulbs is because they get so hot they stay lit between cycles. When you turn off an incad bulb you can even see it slowly shut off for about a second or so till it's fully off.

 

[Jarhead]

Just for the record, I've driven LEDs for periods as short as 4nS (four billionths of a second), and the human eye can see the pulse just fine. However, it is dimmer, since the human eye integrates the light.

If you put a blinking LED in motion, there is no problem seeing separate pulses at 10,000 times a second. However, this does not mean your eyes have a "frame rate" of 10,000Hz.

In my lab, I've also seen that some humans can see things that happen twice as fast, where others don't even detect it. Typically the folks that can see faster things tend to be young, skinny, and usually have much faster metabolisms.

Another interesting item is that the lens in people's eyes actually turns brown with age, and they see the visible spectrum differently.

 

[exdeath]

Frame rate of the eye?

The eye doesn't have a shutter...

Also to add to the examples that the eye is "faster" than people think, you can observe bullets going down range at 900 feet per second.

 

[Six]

Frame rate of the eye?

The eye doesn't have a shutter...

Also to add to the examples that the eye is "faster" than people think, you can observe bullets going down range at 900 feet per second.

I can see black streaks (bullets) from a .22 rifle going down the range, but no one believes me.

 

[Modelworks]

The eye and brain combination in people relies heavily on persistence of vision, what you see in 1 frame of video is blended with the next frame until the previous frame no longer has any effect. A good analogy is using a capacitor with a resistor attached that can drain the capacitor in 2ms, apply different voltages every 1ms and each voltage will carry over to the next reading and the reading you get will not be the current voltage being applied but a combination of previous voltages.

The brain operates at 24Hz, when doing AI experiments with computers that has been the target rate because it is believed that humans can only process 24 'events' per second. The brain takes a ton of shortcuts though so it is able to process information at a much higher perceived rate as long as none of the rules the brain uses are not challenged, optical illusions are examples of when the brains shortcuts fail.

 

[Jeff7181]

The eye and brain combination in people relies heavily on persistence of vision, what you see in 1 frame of video is blended with the next frame until the previous frame no longer has any effect. A good analogy is using a capacitor with a resistor attached that can drain the capacitor in 2ms, apply different voltages every 1ms and each voltage will carry over to the next reading and the reading you get will not be the current voltage being applied but a combination of previous voltages.

The brain operates at 24Hz, when doing AI experiments with computers that has been the target rate because it is believed that humans can only process 24 'events' per second. The brain takes a ton of shortcuts though so it is able to process information at a much higher perceived rate as long as none of the rules the brain uses are not challenged, optical illusions are examples of when the brains shortcuts fail.

This is an odd sentence. You state that as fact, then contradict that by saying "it is believed." Which is it, fact or belief?

 

[Jeff7181]

Just for the record, I've driven LEDs for periods as short as 4nS (four billionths of a second), and the human eye can see the pulse just fine. However, it is dimmer, since the human eye integrates the light.

If you put a blinking LED in motion, there is no problem seeing separate pulses at 10,000 times a second. However, this does not mean your eyes have a "frame rate" of 10,000Hz.

In my lab, I've also seen that some humans can see things that happen twice as fast, where others don't even detect it. Typically the folks that can see faster things tend to be young, skinny, and usually have much faster metabolisms.

Another interesting item is that the lens in people's eyes actually turns brown with age, and they see the visible spectrum differently.

That also has a lot to do with light sensitivity and how our eyes work. Flash a light for 4ns in a dark room and most people will be able to see it. Turn off the lights for 4ns in a bright room and I bet almost nobody will know anything happened.

When I mentioned my experience in a speed reading class, it wasn't just a flash of light. It was black letters on a white/lit background that were displayed at a rate of 500 frames per second.

 

[Smoblikat]

The eye and brain combination in people relies heavily on persistence of vision, what you see in 1 frame of video is blended with the next frame until the previous frame no longer has any effect. A good analogy is using a capacitor with a resistor attached that can drain the capacitor in 2ms, apply different voltages every 1ms and each voltage will carry over to the next reading and the reading you get will not be the current voltage being applied but a combination of previous voltages.

The brain operates at 24Hz, when doing AI experiments with computers that has been the target rate because it is believed that humans can only process 24 'events' per second. The brain takes a ton of shortcuts though so it is able to process information at a much higher perceived rate as long as none of the rules the brain uses are not challenged, optical illusions are examples of when the brains shortcuts fail.

I thought it was 23.967hz

 

[Smoblikat]

The human eye can see in excess of 150fps, have a go with this

That keeps leading me to a wips site.

 

[wuliheron]

The human eye doesn't have a frame rate. It is sensitive enough to detect a single photon and can collect them continuously and send the information immediately to the optic nerve. Exactly how it is processed in the brain is not fully understood at this time.

For display screens 120fps is when the worst of the visual distortions can be eliminated. For example, when a camera pans the background can look blurred because the frame rate is too slow to keep up. If the display itself is not 120fps you get this same kind of visual distortion from it. Some say 180fps or higher for a display screen is about ideal, but 120fps is about as good as you could ask for.

The images you send to the display are another story altogether. Videos are commonly shot at 24fps, while 48fps is considered much better for reducing blurring and whatnot.

Video games are a different animal altogether because they have to react to your input from a mouse or whatever. Human reaction times are around 2ms and about 85fps are considered good enough to eliminate most of the perceptible lag in a game. However, most games are designed to either be played at 30fps or 60fps. That is, the developers add effects like motion blur to hide the fact the game is not as responsive as it could be and can look like a slide show at times.

The slower the game is designed to be played the more such effects are added and reduce the visual quality even if your computer can run the game at 200fps and display it on a 120hz monitor or TV. So for something like a Star Trek holodeck where you are trying to completely fool the human eye with photo realistic images you would need the program to run at least 85fps and be displayed at least 120fps. Some say 180fps or higher display is enough to be indistinguishable from reality.

 

[sm625]

Is there a motion.exe that does 120hz also? Or was that in the link that is now dead?

 

[Modelworks]

I thought it was 23.967hz

I worked as an SFX guy for many years with animation and studied a lot about how the eye perceives color and how the brain processes images and the correct number is 23.976423172 hertz as measured using the latest MRI technology, but 24hz is a lot easier to use in conversation. There is also some who want to include autonomic processes which would greatly change the rate, so like most things it is up for debate.

 

[Modelworks]

This is an odd sentence. You state that as fact, then contradict that by saying "it is believed." Which is it, fact or belief?

I was on the way to work and didn't have time , or really care to, cater to the nitpicking crowd that is internet forums.