Placeholder Image

字幕列表 影片播放

  • Hey it's me Destin.

  • Welcome back to Smarter Every Day.

  • You won't believe your eyes.

  • You've heard this before right?

  • It's usually like a click bait title to get you to watch an internet video

  • or read a stupid article.

  • But are there cases when you actually can't believe your eyes?

  • Make this video as large as you can on the screen that you're watching

  • and we're gonna do an experiment.

  • Put your head at a set distance from the screen and look at this photo.

  • It's a lighthouse on top of an island but I've inverted the colors.

  • I want you to focus your eyes right on the tip of that lighthouse and don't move them.

  • I'm gonna be quiet now and I'm gonna let you transport

  • your mind to this island off the coast of Tasmania.

  • [ seagulls ]

  • OK I'm about to invert the picture back but I want you to stay focused on that lighthouse.

  • Are you ready?

  • Here we go.

  • Out of the corner of your eye you should see a pale blue sky and a deep royal blue sea with little light green grass spots on the island.

  • Now, take your eyes and move it to the edge of the island.

  • You see that? There is no color in this picture.

  • It's a black and white image.

  • Your brain just made up color that wasn't there.

  • Can you believe your eyes?

  • By the way, just in case you're curious

  • this is what the original image looked like.

  • OK, so I just showed you a black and white image

  • and for some reason your brain saw something like this.

  • What is happening?

  • Is that something going on with your eye?

  • Or is that your brain?

  • For all we know it might be the optic nerve connecting the two.

  • Let's investigate a little bit further.

  • A couple of years ago I saw this video on the internet

  • and at first it just looks like a neat little light toy.

  • But my mind saw something way different.

  • I realized for the first time that I literally could not believe my eyes.

  • Not that I didn't understand what was happening,

  • but I knew that what I was seeing didn't actually exist.

  • As soon as I saw this video it was very clear to me that the dude that made this

  • was a genius.

  • The bad news is though,

  • he lives hundreds of miles away in the desert.

  • So I'm in a kind of a sleazy hotel

  • and uh, I don't want to say sleazy.

  • Have you ever seen somebody do something really cool on the internet

  • and you wanted to meet them?

  • Well that's kinda what happened here.

  • This is Greg.

  • - Hello

  • - And Greg has a pretty interesting gadget that you made?

  • Is that what we're gonna call it?

  • - That's what we're gonna call it. The device.

  • - Did you design this PCB?

  • - I did.

  • - Did you populate it?

  • - I did.

  • That's a lot of small chips on there, but..

  • - You're a geek man.

  • - Yeah I know. [laughs]

  • It gets easier.

  • - How long did it take to populate the board?

  • - It takes about an hour and a half.

  • - You did that whole thing in an hour and a half?

  • - Yeah.

  • - Yeah but can you change a water pump on a 1990 Chevrolet pickup?

  • - [laughs]

  • - So those are LEDs right?

  • - Yeah they're red green blue RGB LEDs.

  • - That's pretty cool.

  • Now how are you controlling that?

  • - So these LEDs are mounted on a circuit board that's mounted on a DC motor.

  • And if I apply power to the DC motor, it spins.

  • And if I time it just right, I can essentially light up any LED

  • anywhere on the circle that I want, as we can see here.

  • - So you just created these bitmaps and then uploaded them

  • through some software that you wrote?

  • - Yep, so this is a 63 by 63 pixel bitmap,

  • and essentially I take that and there's actually an

  • infrared sensor on the display that can receive data.

  • - Oh look at that.

  • Smarter Every Day.

  • Little homage.

  • So how did you do that?

  • Did you upload that, or..

  • You did that today didn't you.

  • - Yes actually I just created that image

  • right before driving out to meet you.

  • - This is pretty amazing.

  • So ok, here's the deal.

  • I asked you about this because I wanted to do this.

  • I wanted to use this high speed camera

  • to look at what you've got here.

  • This is a Phantom Miro that we're using here.

  • This is a Miro 320S, and we're gonna setup..

  • What's your update rate on the microcontroller?

  • - Um it's pretty fast. It's spinning at about 25 revolutions every second.

  • - Ahuh.

  • - And within those it updates 256 times.

  • So we're looking at about a 200 microsecond rate that the LEDs get updated.

  • - OK so a thousand frames per second is not fast enough, is what you're telling me.

  • - No.

  • - OK.

  • Alright, well let's figure out what frame rate we need to hit in order to understand what's going on here.

  • - Alright.

  • - What do we call these? Are these pixels?

  • Cause it's not really like a square thing like cartesians,

  • it's kinda like an arc..

  • - Right, it's kind of like an arc pixel.

  • - OK, so what do you call that?

  • - An arxel.

  • - Arxel.. like it. We're gonna go with that.

  • How far does the LED bar travel for each individual arxel as you call it?

  • - It's about one and a half degrees.

  • - If it's only going one to two degrees, then why is my brain still seeing that light

  • the whole time it's around?

  • Because it's only one 360th of the sweep, but the rest of it's dark.

  • According to the high speed camera,

  • which I had to crank up to 5500 frames per second,

  • this is what's actually happening.

  • Check it out.

  • Those's dots, or the arxels as Greg likes to call them

  • are kind of flipping around all over the place.

  • Most of the image is dead space.

  • So why is your brain making that image?

  • To understand why the brain sees something that's not there

  • I found a guy that studies this sort of thing

  • that's published over 130 different papers on similar topics.

  • OK I'm on a pretty bad Skype connection with Dr Stuart Anstis

  • who is a genius at the University of California San Diego

  • and what do you study Dr Anstis?

  • - [british accent] I study visual perception, in particular visual illusions

  • which tell us about the normal processes of vision,

  • how the eyes send information to the brain.

  • - That's fantastic. And obviously your accent makes it very clear

  • that you know exactly what you're talking about because I would

  • expect nothing less from a person who studies visual perception.

  • So I want to understand why my eyeball is seeing something where I know there is not light.

  • Why am I seeing that?

  • - It's because of persistence of vision, which means the eye averages what it sees over a short period of time.

  • It's analogous to a camera where you have a long exposure time,

  • and this will give you more light coming in,

  • greater sensitivity, but you have a more sluggish response.

  • So anything moving gets blurred out.

  • - What is the difference in time from the moment the LED is illuminated

  • until my eye registers that the light is there?

  • There has to be a delay time there.

  • What is that?

  • - That delay time varies enormously, over a ten-fold range,

  • anything from 10 to 100 milliseconds.

  • - According to Dr Anstis there's two things going on,

  • and let's look at it.

  • Let's pretend that we have a flashing LED

  • and we want to look at the brain's response to that LED.

  • First of all he said there's a delay,

  • so when the LED first comes on, our brain's not going to

  • immediately see vision, it's gonna take some finite amount of time later.

  • Secondly he said that the eye averages what it sees over a short period of time.

  • Think about that.

  • If we have a moving average, that means that our vision has some sort of inertia to it.

  • It works like this.

  • As the average comes along and is exposed to that LED flash, it starts to ramp up.

  • As the light goes away, that moving average starts to ramp back down.

  • As the light comes back that average starts to go up again,

  • and instead of having gaps that are complete darkness, we have this nice trough in the bottom.

  • Therefore we have a persistence of vision even though there is no light to see at that point in time.

  • Let's look back at that slo mo image from before with all the blinking LEDs.

  • Now, let's add this time average of light and see what the image looks like.

  • Check it out.

  • How cool is that?

  • It looks just like what our eyes see.

  • At what rate would you expect that I would quit seeing a uniform image

  • but I would start to see like a tail dragging across the screen?

  • - Well if the average time is less than one revolution then you're going to see a gap.

  • But supposing as you say, the propellor goes around in 1/25th of a second, that's 40 milliseconds.

  • - I get it now.

  • Do you understand?

  • Think about it.

  • Greg's wheel is rotating at 25 frames per second,

  • and that has to do with the moving time average of the human eye.

  • That's why this video is at least 25,

  • it's actually 30 frames per second.

  • If it wasn't, you would see flickering of the image.

  • So what seems like an imperfection in our eye

  • is actually what smooths things out and makes the world work smoothly for us.

  • That's pretty awesome.

  • If the people watching this video were students in your class

  • what would you want them to know about Greg's wheel and the persistence of vision?

  • - [laughs] Well I would say every system has got a limited time resolution

  • The eye has an engineering problem of trading time resolution against sensitivity to light.

  • And in fact it's got sort of knobs inside which can change that trade-off relationship automatically.

  • The eye is in many ways much cleverer than the camera.

  • It's a beautiful piece of engineering.

  • - Well thank you very much sir, I really appreciate your time.

  • - Thank you.

  • - Alright I hope you enjoyed this episode of Smarter Every Day.

  • It was sponsored by lynda.com which helps pay for people like Micah who's a video editor.

  • - Hello.

  • - Who happens to actually have a certification from Lynda.

  • Lynda is like a Smarter Every Day but on very technical topics.

  • So in this particular case, after effects.

  • I had no idea how to do this.

  • So Micah came along and helped me.

  • He actually took a bunch of classes on Lynda and we found one together right?

  • - That's right.

  • - And what was it called?

  • - Echo.

  • - This was the tutorial that taught me to model persistence of vision using After Effects.

  • But you can learn almost any topic you want.

  • They've got Excel tutorials, Photoshop tutorials, even how to edit videos like I do with Premiere.

  • This is not the stuff you're gonna find on YouTube, this is really high quality tutorials.

  • They've got over 100,000 of them.

  • So if you want to support Smarter Every Day,

  • please consider going to lynda.com/smarter

  • You'll get a 10 day free trial which you can cancel at any time.

  • I wanted to make sure that if you're supporting Smarter Every Day

  • you're gonna have a really good experience

  • so that's lynda.com/smarter

  • You can cancel at any time.

  • I'm Destin,

  • I hope you enjoyed this, feel free to subscribe,

  • if not, no big deal.

  • Hope you're getting Smarter Every Day

  • Have a good one.

  • OK is there any truth to the rumor that you actually proposed using one of these?

  • - There is some truth. [laughs]

  • As I was finishing up the project I thought to myself,

  • Hey it'd be pretty cool if I could like write messages on here

  • so I decided that's how I wanted to propose.

Hey it's me Destin.

字幕與單字

單字即點即查 點擊單字可以查詢單字解釋

A2 初級

"你不會相信你的眼睛!"- "每天更聰明 "142 ("YOU WON'T BELIEVE YOUR EYES!" - Smarter Every Day 142)

  • 110 10
    SylviaQQ 發佈於 2021 年 01 月 14 日
影片單字