4 ms·

There are many examples of humans being able to observe a difference, yet deny their capability to discern on the basis of abstract rational knowledge. Another

There are many examples of humans being able to observe a difference, yet deny their capability to discern on the basis of abstract rational knowledge. Another example is frame rates: while its well known that the bandwidth of light sensitive cones on the retina is around 30Hz cycle frequency (and thus ~60Hz framerate), many gamers felt the difference of higher framerates, even though many of the same gamers denied observability of this difference on the basis of knowledge of this basic biological fact. The biological fact is not wrong: if screen pixel projects are at rest on the retina. But when the eye is rotating to follow a depicted object on the screen, then speeds higher than ~60 pixels per second result in motion blur that would be absent when observing a real life object moving at the same angular speeds. The biological fact is still true, but simply not naively applicable in the context of motion blur due to eyeballs rotating while observing a display.

ajuc 2y ago
It's more complicated than that, brain cuts out the blurry "in between" frames when you move your eyes. Which means it has to extrapolate the images from before the movement to fill the gaps. There's many weird effects of this hack - for example when you look at something moving very fast (like wheels of passing vehicles) but don't move your eyes - it looks blurry. But when you move your eyes - the frame in your brain "freezes" and you can notice non-blurry details on the fast moving things. There are even ways to force eyes to see stuff from the past (like seconds' hand on a clock moving at different rates depending if you move your eyes). You can't actually say that human eye has a strict fps, so the higher screen fps the less weird interactions with the ugly hacks in our brains.
- DoctorOetker 2y ago
  when your eyes are motion tracking an object there is no cutting-out going on, that only happens when the brain intentionally decides to track a new object
Kudos 2y ago
Your eyes being analog and the 60Hz display being offset from your own cycle frequency would play large roles in that.
- DoctorOetker 2y ago
  the phenomenon I describe is much simpler, and is remediated by having the backlight of a 60 Hz display have a very short single on strobe. If they backlight remains on or flashes multiple times during a single frame then the pixel will draw a long streak on the retina (because the eye was rotating to track a depicted moving object)
  - thfuran 2y ago
    Sample and hold motion blur during smooth pursuit is far from the only visual artifact resulting from low frame rates.
    - wizzwizz4 2y ago
      What are others?
      - thfuran 2y ago
        I'm not sure if you've seen them, but there are LED glow sticks that look normal when stationary but leave sort of a trail of distinct after images rather than a smooth blur when waved. I've seen LED Christmas lights that have a similar effect when you scan your eyes past them quickly. It's called the phantom array effect, and it's caused by a flickering light (LEDs lit with less than 100% duty cycle) moving rapidly across the visual field. Mitigating sample and hold motion blur by blacking out the display for some of the frame time turns your display into a flickering light source, so it can potentially produce the same effect for small bright objects on dark backgrounds when either the object moves rapidly on screen or you scan your eyes across the screen. It's fairly niche in that it'd only affect some rather specific scenes and even then can depend on how the scene is viewed (it won't occur if you track the only fast-moving bright object with your eyes, and it's much reduced in higher ambient light), but the flicker fusion rate is generally several kilohertz, so you'd need to boost the hell out of the frame rate to guarantee the effect is eliminated altogether. More generally, an object that is moving rapidly enough for its screen position to change by many pixels per frame can have its motion look jerky even without motion blur. And, in fact, motion blur can help disguise that.
Ferret7446 2y ago
I think you're the one denying rational knowledge here. People are different, and there are in fact people who cannot tell the difference between higher frame rates, as shown in studies. Just because you can see it doesn't mean others can.
- DoctorOetker 2y ago
  the problem is not rational knowledge, its misapplying rational knowledge in circumstances where it does not apply. its extremely easy to reproduce, and I have not encountered anyone who fails to see it once told how to reproduce it: suppose your screen is 1920 pixels wide (Full HD). suppose a patter contains high spatial bandwidths, i.e. high spatial resolutions like text. obviously if a monitor is able to display say at a 60 Hz frame rate, then motion blur would be essentially absent if the font bitmap is moved 1 pixel per frame, or at least the motion blur would be constrained to 1 pixel of horizontal blur. to traverse about 1920 pixels from the left side of the screen to the right at 1 pixel per frame (or 60 pixels per second) would take 32 SECONDS. Half a minute! consider much faster velocities, say 3 seconds for a word to move from the left side of your screen to the right, thats motion blur such that a pixel is blurred the length of 10 pixels! Most people would not consider this very fast at all. now take note of the DPI of the screen and print the same text at the same dimensions on paper, and hold it up close to your screen, and move it from left to right in about 3 seconds, while following it with your eye. You wont see motion blur on the real piece of paper, but you will on the digitally rendered text of same dimensions and velocity... once you understand this phenomenon, you will understand in what sense the flicker bandwidth of human retina at rest with respect to a visual stimulus should not be confused for a sufficient framerate to accurately reproduce the perception of visual stimuli representing motion, unless the backlight is strobed at a very low duty cycle. but then flicker becomes visible again unless we bump the framerate up a notch again.

ajuc 2y ago

It's more complicated than that, brain cuts out the blurry "in between" frames when you move your eyes. Which means it has to extrapolate the images from before the movement to fill the gaps. There's many weird effects of this hack - for example when you look at something moving very fast (like wheels of passing vehicles) but don't move your eyes - it looks blurry. But when you move your eyes - the frame in your brain "freezes" and you can notice non-blurry details on the fast moving things. There are even ways to force eyes to see stuff from the past (like seconds' hand on a clock moving at different rates depending if you move your eyes). You can't actually say that human eye has a strict fps, so the higher screen fps the less weird interactions with the ugly hacks in our brains.

DoctorOetker 2y ago
when your eyes are motion tracking an object there is no cutting-out going on, that only happens when the brain intentionally decides to track a new object

Kudos 2y ago

Your eyes being analog and the 60Hz display being offset from your own cycle frequency would play large roles in that.

DoctorOetker 2y ago
the phenomenon I describe is much simpler, and is remediated by having the backlight of a 60 Hz display have a very short single on strobe. If they backlight remains on or flashes multiple times during a single frame then the pixel will draw a long streak on the retina (because the eye was rotating to track a depicted moving object)
- thfuran 2y ago
  Sample and hold motion blur during smooth pursuit is far from the only visual artifact resulting from low frame rates.
  - wizzwizz4 2y ago
    What are others?
    - thfuran 2y ago
      I'm not sure if you've seen them, but there are LED glow sticks that look normal when stationary but leave sort of a trail of distinct after images rather than a smooth blur when waved. I've seen LED Christmas lights that have a similar effect when you scan your eyes past them quickly. It's called the phantom array effect, and it's caused by a flickering light (LEDs lit with less than 100% duty cycle) moving rapidly across the visual field. Mitigating sample and hold motion blur by blacking out the display for some of the frame time turns your display into a flickering light source, so it can potentially produce the same effect for small bright objects on dark backgrounds when either the object moves rapidly on screen or you scan your eyes across the screen. It's fairly niche in that it'd only affect some rather specific scenes and even then can depend on how the scene is viewed (it won't occur if you track the only fast-moving bright object with your eyes, and it's much reduced in higher ambient light), but the flicker fusion rate is generally several kilohertz, so you'd need to boost the hell out of the frame rate to guarantee the effect is eliminated altogether. More generally, an object that is moving rapidly enough for its screen position to change by many pixels per frame can have its motion look jerky even without motion blur. And, in fact, motion blur can help disguise that.

Ferret7446 2y ago

I think you're the one denying rational knowledge here. People are different, and there are in fact people who cannot tell the difference between higher frame rates, as shown in studies. Just because you can see it doesn't mean others can.

DoctorOetker 2y ago
the problem is not rational knowledge, its misapplying rational knowledge in circumstances where it does not apply. its extremely easy to reproduce, and I have not encountered anyone who fails to see it once told how to reproduce it: suppose your screen is 1920 pixels wide (Full HD). suppose a patter contains high spatial bandwidths, i.e. high spatial resolutions like text. obviously if a monitor is able to display say at a 60 Hz frame rate, then motion blur would be essentially absent if the font bitmap is moved 1 pixel per frame, or at least the motion blur would be constrained to 1 pixel of horizontal blur. to traverse about 1920 pixels from the left side of the screen to the right at 1 pixel per frame (or 60 pixels per second) would take 32 SECONDS. Half a minute! consider much faster velocities, say 3 seconds for a word to move from the left side of your screen to the right, thats motion blur such that a pixel is blurred the length of 10 pixels! Most people would not consider this very fast at all. now take note of the DPI of the screen and print the same text at the same dimensions on paper, and hold it up close to your screen, and move it from left to right in about 3 seconds, while following it with your eye. You wont see motion blur on the real piece of paper, but you will on the digitally rendered text of same dimensions and velocity... once you understand this phenomenon, you will understand in what sense the flicker bandwidth of human retina at rest with respect to a visual stimulus should not be confused for a sufficient framerate to accurately reproduce the perception of visual stimuli representing motion, unless the backlight is strobed at a very low duty cycle. but then flicker becomes visible again unless we bump the framerate up a notch again.