[00:01.330 --> 00:02.230]  Thank you, everybody.
[00:02.470 --> 00:03.010]  Thanks for coming.
[00:03.190 --> 00:04.290]  I'm stoked to be here.
[00:05.830 --> 00:06.810]  Let's get started.
[00:06.990 --> 00:07.890]  Great lights are off.
[00:08.550 --> 00:09.090]  All right.
[00:09.390 --> 00:10.630]  Light, color, and perception.
[00:10.790 --> 00:19.270]  This is going to be a great talk to start out the weekend with, because you've got lots of pretty colors and not too much hexadecimal, so you can kind of ease into the weekend.
[00:20.350 --> 00:20.770]  All right.
[00:23.150 --> 00:27.350]  Start with a presentation with a quote from the Dalai Lama.
[00:27.670 --> 00:28.790]  Yeah, I'm from California.
[00:30.570 --> 00:36.150]  I think the Dalai Lama is a heavy cat, and I think he's kind of a hacker, right?
[00:36.290 --> 00:41.770]  Because he's talking about learning the system so you can hack it properly.
[00:42.270 --> 00:49.770]  So the system I'm going to be talking about today is kind of the visual system, your perceptual system, and the physics between light and color.
[00:50.490 --> 00:56.090]  I don't have any particular qualifications to talk about this other than it fascinates me.
[00:56.090 --> 00:59.630]  I have a PhD, but not in horribleness or any of this stuff.
[01:01.070 --> 01:10.530]  I do make and sell some of these little perceptual hacks, which I tell people is art, and for some reason that, I think, makes me a professional.
[01:11.410 --> 01:11.510]  Okay.
[01:11.730 --> 01:13.730]  So, enough about me.
[01:13.950 --> 01:16.150]  Let's talk about you, okay?
[01:16.310 --> 01:18.310]  Specifically what's inside your eyeball.
[01:18.310 --> 01:28.610]  And the way you see the world is through these specific specialized cells called rods and cones, and they're the back of your retina, and the retina is the back of your eyeball.
[01:29.530 --> 01:30.050]  Okay.
[01:30.070 --> 01:32.070]  And here's kind of the structure of things.
[01:32.450 --> 01:34.050]  Let's see if I got a pointer here.
[01:34.530 --> 01:34.730]  Yeah.
[01:34.970 --> 01:44.370]  So, this is the light comes in from your left here, and the sensors are actually behind all the structure of the wiring and the plumbing or something.
[01:44.370 --> 01:53.210]  So, if you ever need a handy argument for evolution, you know, if you were going to design this thing from scratch, you'd probably want to put the sensors in front of the wiring and the plumbing.
[01:55.190 --> 01:58.090]  So, looks like my fonts got messed up.
[01:58.510 --> 02:08.230]  So, anyway, there's two kinds of, actually four kinds, but two major kinds of cells, light sensor cells.
[02:08.230 --> 02:11.310]  One is the rod, and those are basically brightness.
[02:11.570 --> 02:12.830]  They're not, they don't detect color.
[02:14.730 --> 02:21.530]  They're the ones that work when you have night vision, when it's too dim to see color, you know, when it's super dark.
[02:22.390 --> 02:28.330]  And the ones that I'm going to be talking most about are these things called cones, which are specific color sensors.
[02:30.210 --> 02:34.250]  And they're active, they're a lot more precise, they're more acute.
[02:35.510 --> 02:43.150]  If you know what the fovea is, that's the part in the center of your eye, which is really the most sensitive and acute.
[02:44.670 --> 02:48.010]  Okay, so, I have no idea if this is going to work.
[02:48.190 --> 02:50.570]  This is a little demonstration.
[02:51.390 --> 02:57.850]  If you could, this may work better, the closer you are to the screen, if you could stare at that red dot for a couple seconds.
[02:58.450 --> 03:02.010]  All right, just try not to move your eyes, even though that's impossible, because your eyes move naturally.
[03:02.730 --> 03:06.010]  I apologize for people who can't see this straight on.
[03:06.470 --> 03:13.270]  And if you stare at that red dot without moving your eyes, you might notice something that's happening to that green circle.
[03:13.710 --> 03:15.090]  Is anybody noticing anything?
[03:15.250 --> 03:17.570]  Like, maybe it's, maybe it's going away?
[03:18.730 --> 03:19.470]  Right, right.
[03:19.710 --> 03:20.370]  Okay, yeah.
[03:20.550 --> 03:24.890]  So, if this is working for you, and I apologize if it isn't, this is called Troxler fading.
[03:25.150 --> 03:30.130]  Basically, what you're doing is you're kind of super saturating the green detectors in your eye.
[03:30.310 --> 03:32.050]  And so, that will kind of disappear.
[03:32.370 --> 03:33.730]  Now, I'm going to show a white background.
[03:34.970 --> 03:38.450]  And if you look right in the middle of that, does anybody see an after image?
[03:39.030 --> 03:39.850]  Yeah, okay.
[03:40.050 --> 03:41.010]  What color was that?
[03:41.850 --> 03:42.250]  Reddish.
[03:42.250 --> 03:44.530]  Kind of reddish, pinkish, purplish something.
[03:44.850 --> 03:45.130]  All right.
[03:45.130 --> 03:48.490]  Okay, we can explain all this stuff from physics and physiology.
[03:48.770 --> 03:50.650]  And that's mostly what this talk is about.
[03:50.850 --> 03:57.790]  Kind of how your perceptual system works, and how colors work, and how things get fooled.
[03:57.950 --> 03:58.750]  Which is kind of cool.
[03:59.230 --> 03:59.450]  All right.
[03:59.590 --> 04:02.910]  So, we talked about the three kind of cone sensors.
[04:04.450 --> 04:11.630]  There's one that's dedicated to blue, and there's one kind of green, and then there's one which is orange-red.
[04:11.630 --> 04:14.430]  And they're not precise, and you can see there's a lot of overlap.
[04:14.650 --> 04:16.150]  But this is how you see colors.
[04:17.330 --> 04:17.770]  Okay.
[04:18.790 --> 04:26.870]  And so, you can see the entire visible spectrum, basically, through these three sensors, which are sensible to these particular wavelengths.
[04:27.090 --> 04:35.730]  And basically, if you look at a particular color, it will give you a response on these different sensors in a different way, which will give you the sensation of color.
[04:35.730 --> 04:47.290]  And we refer to color with these various names, and if you're into the physics of it, you can also talk about the wavelength of the light that produces the color.
[04:47.510 --> 04:52.110]  So, you know, it's an electromagnetic oscillation, and it has a particular wavelength.
[04:52.110 --> 04:55.950]  And 400 nanometers for violet, 550 is green.
[04:56.170 --> 05:00.290]  That's more or less the center of the visual spectrum, and goes to infrared.
[05:00.650 --> 05:13.710]  And if you're a ham, you know it goes all the way into RF, and then down into the RF, and all the way up into gamma rays and X-rays, and stuff like that.
[05:13.850 --> 05:18.450]  But this is just one tiny little fraction of the electromagnetic spectrum that we can actually see.
[05:18.450 --> 05:31.350]  There's some interesting stuff, and it turns out that the lower bound of the rod, sorry, cone cells that detect blue, it turns out not to be the cone cells.
[05:31.490 --> 05:38.470]  They can detect blue, further deeper blue, into the ultraviolet than you can see.
[05:38.530 --> 05:45.550]  And what's actually happening is that the lens and the cornea blocks out some of that ultraviolet.
[05:45.550 --> 05:50.150]  So apparently, people, you can get an operation that replaces your lens if you have cataracts.
[05:50.350 --> 05:53.830]  And when that happens, you can see, apparently, ultraviolet.
[05:53.870 --> 05:58.310]  You can see further into the purple than you could normally, which I think is kind of cool.
[05:58.550 --> 06:04.530]  So maybe that'll be the next thing for hacking and body modding to get the lens replacements.
[06:05.430 --> 06:05.990]  All right.
[06:06.530 --> 06:09.950]  So, you talk about kind of full-spectrum color.
[06:10.210 --> 06:18.850]  And if you've ever... I hope you've done this experiment where you take the white light, and you put it into a prism, and you see the beautiful rainbow.
[06:19.230 --> 06:19.490]  Whoops.
[06:21.210 --> 06:22.870]  And that's actually a pretty thing.
[06:23.010 --> 06:27.130]  You've all seen the video where this guy is crying about it just because it's so awesome, right?
[06:28.390 --> 06:30.050]  Yeah, so it's awesome.
[06:30.470 --> 06:30.570]  Rainbow.
[06:31.030 --> 06:31.430]  Exactly.
[06:33.270 --> 06:37.270]  So, this is the rainbow, and this has all the colors we can see, right?
[06:37.870 --> 06:39.230]  Well, it turns out it doesn't.
[06:39.410 --> 06:42.890]  Okay, there's this kind of pinkish-purplish, technically magenta.
[06:43.790 --> 06:45.750]  And, you know, where does that come from?
[06:45.830 --> 06:48.270]  That's not actually in the spectrum, this purple color.
[06:48.550 --> 06:53.810]  And what that turns out to be, that turns out to be a mix of blue-red, blue light and red light.
[06:53.810 --> 06:58.090]  And you put them together and you get this magenta thing, but it's not actually in the spectrum.
[06:58.250 --> 07:04.850]  There's no wavelength of light that has this particular color, which is kind of neat.
[07:05.710 --> 07:06.170]  Alright.
[07:06.770 --> 07:09.590]  Now, the way this all works is you can mix colors.
[07:10.030 --> 07:13.910]  You can mix colors light just by adding them together, just by shining them.
[07:14.270 --> 07:20.750]  And this shows how you can get, basically, the primary colors are red, green, and blue.
[07:20.750 --> 07:24.590]  And if you mix those in certain proportions, you can get every color of the spectrum.
[07:24.710 --> 07:25.750]  Every color you can see.
[07:25.890 --> 07:30.190]  And this is very much related to the fact that we have these three color-sensing cones.
[07:31.930 --> 07:34.890]  And if you mix them all in the right proportion, you get white.
[07:35.010 --> 07:40.130]  So white is basically the sum total of all the colors all mixed together.
[07:40.670 --> 07:41.110]  Alright.
[07:41.670 --> 07:55.390]  And this is a really handy thing that we can generate all these colors from three basic primary colors, because that means we can make these nice displays, we can make these nice projectors, these little computers that we keep in our pants.
[07:55.390 --> 08:04.650]  Now we can see these super nice colors by basically lighting up or darkening these three little elements.
[08:04.750 --> 08:06.550]  Here's a LCD display.
[08:06.890 --> 08:10.130]  I hope everyone has looked closely at your computer screen.
[08:10.190 --> 08:11.870]  You can see that it's made up these little dots.
[08:12.050 --> 08:14.170]  There's a red dot, a green dot, and a blue dot.
[08:14.170 --> 08:19.810]  And if you light those up in different brightnesses, you get the perception of color.
[08:20.490 --> 08:20.970]  Okay.
[08:21.130 --> 08:23.730]  So this is the RGB color cube.
[08:23.970 --> 08:28.070]  And typically you can get any color you want just by mixing RGB.
[08:28.710 --> 08:33.130]  And if you have all red and no green and no blue, then you get red.
[08:33.130 --> 08:35.590]  If you mix red and blue, you get magenta.
[08:35.590 --> 08:38.150]  Like we said, if you mix blue and green, you get cyan.
[08:38.150 --> 08:46.550]  And this has basically the entire rainbow in here and even the colors that you can't see, that aren't in the rainbow like magenta.
[08:47.570 --> 08:48.090]  Alright.
[08:48.590 --> 08:49.830]  A little digression.
[08:50.250 --> 08:55.670]  Some of you, some of you I hope are fans of this comic by Randall Munro.
[08:55.930 --> 09:04.930]  He did the color survey where basically he put up a million different colors based on the web and asked people to name what the color was.
[09:04.930 --> 09:09.310]  Because he just generated color at random and said, hey, what do you call this color?
[09:09.470 --> 09:12.750]  So if the color was something around here, most people called it green.
[09:13.290 --> 09:20.630]  And you can see how it goes from teal to blue and the colors that you might expect.
[09:21.950 --> 09:28.730]  It had a really interesting thing because in the survey he took other information like gender.
[09:28.930 --> 09:32.090]  He could figure out some interesting gender differences.
[09:32.090 --> 09:39.750]  And these are the color names that are disproportionately used by women instead of men.
[09:39.970 --> 09:45.190]  So these are color names that women used a lot but men used very rarely.
[09:45.590 --> 09:45.770]  Okay.
[09:46.470 --> 09:50.570]  So I just maybe think for a moment what...
[09:50.570 --> 09:52.450]  I'm going to show you what the guy said.
[09:52.710 --> 09:56.590]  And think for a moment, guess what the most popular color name is for a guy.
[10:06.720 --> 10:07.120]  Okay.
[10:07.380 --> 10:07.740]  Right.
[10:07.900 --> 10:07.980]  Right.
[10:08.100 --> 10:08.140]  Right.
[10:08.360 --> 10:08.580]  Right.
[10:08.760 --> 10:09.160]  Okay.
[10:09.260 --> 10:11.040]  This is absolute real data, right?
[10:11.200 --> 10:15.140]  Even the misspelling of number five, that's how it was typed in.
[10:15.720 --> 10:17.720]  So go team Y chromosome.
[10:18.080 --> 10:18.180]  Right.
[10:18.340 --> 10:18.540]  We're good.
[10:18.740 --> 10:19.000]  All right.
[10:20.500 --> 10:20.900]  Okay.
[10:21.020 --> 10:25.540]  So he has all the data if you want to do these experiments yourself, if you're a sociologist or something.
[10:25.540 --> 10:28.020]  I think this is pretty fascinating stuff.
[10:28.020 --> 10:28.800]  Okay.
[10:29.080 --> 10:30.120]  So where were we?
[10:31.340 --> 10:31.780]  Okay.
[10:31.920 --> 10:33.120]  So we were talking about light.
[10:33.340 --> 10:36.340]  And you can add light just by shining different colors of light.
[10:36.540 --> 10:38.140]  But not everything is light.
[10:38.860 --> 10:46.980]  If you're doing printed things, if you're an artist, if you're using colors, if you're using pigments, what you want to do is use subtractive colors.
[10:47.300 --> 10:51.660]  And the subtractive colors are yellow, magenta, and cyan.
[10:52.360 --> 10:58.060]  And basically, the way those work is they basically filter out stuff from light.
[10:58.200 --> 11:05.720]  So if you take white and you put it through a yellow filter, it basically removes the blue.
[11:06.180 --> 11:13.200]  So here's kind of an example of camera response.
[11:13.320 --> 11:14.640]  Oops, that slide is in the wrong place.
[11:14.640 --> 11:22.240]  But when you put it all together, you can kind of make this color wheel.
[11:22.460 --> 11:27.100]  And you have the primary colors of light, which are red, green, blue.
[11:27.300 --> 11:29.580]  And then there are these kind of complementary colors.
[11:30.040 --> 11:32.660]  So yellow is basically...
[11:32.660 --> 11:33.360]  Oops.
[11:34.780 --> 11:37.340]  Yellow is red plus green.
[11:37.480 --> 11:38.780]  Now this doesn't sound right.
[11:38.780 --> 11:43.440]  If you're used to like Crayola crayons, you may remember something.
[11:44.620 --> 11:46.300]  But it's actually true in light.
[11:46.520 --> 11:55.900]  And if you see some of the LED displays, which have red, green, and this kind of orangey-yellow thing, that's just made by mixing the red and the green in the LED.
[11:57.920 --> 12:01.460]  So basically, you can take white and subtract a color.
[12:01.620 --> 12:07.120]  If you subtract red from white, you get this magenta thing.
[12:07.340 --> 12:08.680]  I'm sorry, that's wrong.
[12:08.880 --> 12:11.260]  If you subtract green from white, you get magenta.
[12:11.260 --> 12:13.360]  And that explains the green after image.
[12:13.780 --> 12:14.380]  Okay.
[12:15.640 --> 12:20.100]  Here's an interesting hack that people did, a designer did.
[12:20.200 --> 12:26.060]  This is wallpaper that changes basically aspect depending on what color you illuminate it with.
[12:26.200 --> 12:33.740]  So there's basically three pictures here kind of on top of each other printed in cyan, magenta, and yellow.
[12:36.140 --> 12:41.520]  And if you illuminate them in different colors, you get to see the different images.
[12:41.560 --> 12:45.760]  For example, on the left there, the bones are in red.
[12:45.940 --> 12:49.740]  In the actual color, the bones are in cyan.
[12:50.060 --> 12:52.400]  So cyan is a complementary color of red.
[12:52.400 --> 12:54.000]  So that's going to absorb the red.
[12:54.000 --> 12:56.420]  So that looks dark and everything else looks bright.
[12:56.680 --> 13:06.220]  So this is kind of a nice hack how you can get three different images from the three different colors using reflection.
[13:06.620 --> 13:10.080]  And it's a nice demonstration of the complementary colors.
[13:10.620 --> 13:11.120]  Okay.
[13:11.320 --> 13:15.360]  So here's another little demonstration.
[13:16.120 --> 13:20.260]  If you want to stare at that red dot, the red star for a while.
[13:20.560 --> 13:22.660]  And the longer you stare at it, the better this works.
[13:22.920 --> 13:25.400]  So I'm going to kill some time here and talk about this.
[13:26.240 --> 13:29.260]  Yeah, I'm from California and you might recognize the picture there.
[13:30.060 --> 13:34.460]  This is another demonstration of complementary colors.
[13:36.420 --> 13:38.740]  And just give that a couple more minutes.
[13:38.960 --> 13:40.160]  Try not to move your eyes.
[13:40.300 --> 13:42.560]  Look at that red dot if you can.
[13:42.560 --> 13:46.100]  And now I'm going to show that in a little higher res.
[13:46.800 --> 13:46.980]  Okay.
[13:47.880 --> 13:48.980]  And you will notice...
[13:48.980 --> 13:50.740]  Anybody see a color image there for a minute?
[13:50.920 --> 13:51.180]  Yes.
[13:51.380 --> 13:51.620]  Right.
[13:51.740 --> 13:52.840]  It's not a color image.
[13:52.940 --> 13:53.560]  It's black and white.
[13:53.800 --> 13:54.260]  Right.
[13:57.970 --> 13:58.450]  Okay.
[13:58.630 --> 14:00.730]  So that's called the Spanish castle effect.
[14:00.730 --> 14:03.670]  And that's basically the same effect as the...
[14:03.670 --> 14:04.350]  as the...
[14:04.350 --> 14:05.170]  as the tracts are fading.
[14:05.690 --> 14:13.130]  When you look at this particular orangey thing, you kind of saturate the...
[14:13.130 --> 14:13.550]  the...
[14:14.130 --> 14:14.470]  the...
[14:14.470 --> 14:15.050]  the...
[14:15.050 --> 14:17.350]  the red and green receptors in your eye.
[14:17.550 --> 14:22.210]  And when you show it basically this gray color, it then looks bluish.
[14:23.330 --> 14:23.890]  So...
[14:23.890 --> 14:25.670]  that's kind of a neat effect.
[14:26.210 --> 14:26.530]  All right.
[14:27.810 --> 14:28.230]  Okay.
[14:28.430 --> 14:28.810]  Is it...
[14:29.630 --> 14:30.190]  Okay.
[14:30.470 --> 14:31.510]  Back to color spaces.
[14:31.710 --> 14:31.890]  All right.
[14:32.110 --> 14:33.090]  If you're...
[14:33.090 --> 14:39.030]  if you're a computer programmer, you've ever worked with image processing, the color spaces are pretty natural.
[14:39.210 --> 14:41.030]  You can represent any color with this...
[14:41.030 --> 14:43.450]  this triplet of red, green, and blue.
[14:43.670 --> 14:43.770]  Right.
[14:43.950 --> 14:44.310]  And you...
[14:44.310 --> 14:45.370]  you give it all red.
[14:47.770 --> 14:49.370]  You make it red and things like that.
[14:50.010 --> 14:50.410]  And...
[14:50.410 --> 14:54.650]  if you're a mathematician, you can think of this as kind of this abstract space.
[14:54.650 --> 14:55.650]  It has three dimensions.
[14:55.830 --> 14:56.890]  It has the red dimension.
[14:57.130 --> 14:58.250]  It has the green dimension.
[14:58.310 --> 14:59.270]  It has the blue dimension.
[14:59.450 --> 15:02.810]  So any particular color is just a particular point in this space.
[15:02.810 --> 15:05.950]  And you can give that a numerical number.
[15:06.590 --> 15:08.190]  And that describes a color.
[15:08.330 --> 15:14.570]  So that's a nice way to describe a color in an image format, is just this triplet of numbers.
[15:15.590 --> 15:16.190]  Okay.
[15:16.290 --> 15:18.450]  There's other ways to describe colors.
[15:18.670 --> 15:20.130]  Hue, saturation, value.
[15:20.610 --> 15:22.030]  Hue, saturation, lightness.
[15:22.110 --> 15:31.510]  And this way, there's a number which is an angle basically around this cylinder, which gives you the hue, which is basically the color.
[15:32.150 --> 15:35.650]  Then there's a saturation, which is how much color it is.
[15:35.810 --> 15:37.550]  If it's unsaturated, it's pastel.
[15:37.690 --> 15:39.970]  If it's completely unsaturated, it's white.
[15:40.190 --> 15:42.490]  And the value is basically how bright it is.
[15:42.570 --> 15:44.670]  If it's a pure color or if it's black down here.
[15:44.750 --> 15:49.150]  There's also hue, saturation, lightness, which is kind of like this, but slightly different.
[15:49.350 --> 15:51.010]  I won't go into it and bore you to death.
[15:51.990 --> 15:52.430]  Okay.
[15:52.550 --> 15:59.330]  So if you look at things that are fully saturated, you can get some interesting effects like this.
[15:59.330 --> 16:02.590]  And I apologize if you're a little hungover because this may...
[16:04.810 --> 16:08.950]  This is kind of... this kind of pops and vibrates and stuff like that.
[16:09.070 --> 16:14.290]  I'm told that if you're a designer, which I'm not, you're not supposed to use stuff like this.
[16:14.390 --> 16:24.210]  And what this is, is this is basically two colors that are fully bright, but they're basically... they have no shade difference.
[16:24.210 --> 16:25.950]  They're the same tone of brightness.
[16:26.510 --> 16:33.230]  And your visual system uses edges to detect things and it looks for changes in brightness.
[16:33.410 --> 16:35.470]  Now these colors are exactly the same bright.
[16:35.630 --> 16:37.490]  They're just different colors, right?
[16:37.710 --> 16:41.670]  So your visual system doesn't have a... doesn't have much to grab onto.
[16:41.830 --> 16:43.030]  And so it just kind of vibrates.
[16:43.170 --> 16:46.330]  I was looking for a better explanation of that, but I couldn't really find one.
[16:46.470 --> 16:48.090]  So you'll have to do this.
[16:48.090 --> 16:50.510]  And there's also some nice illusions out there.
[16:50.630 --> 16:55.070]  There's this Japanese fellow named Kedaoka who is... just does these stunning things.
[16:55.230 --> 16:56.130]  And I don't know if you can see that.
[16:56.230 --> 16:57.010]  It's kind of moving.
[16:57.330 --> 17:01.710]  I was really looking for a good explanation as to why this happens because I'd love to explain it.
[17:01.810 --> 17:03.410]  But I'm not sure anybody really knows.
[17:03.410 --> 17:10.650]  There's all sorts of things going on in your retina and in the lower parts of the vision system of your brain.
[17:10.970 --> 17:15.290]  And going up all the way to your visual centers.
[17:16.770 --> 17:17.250]  Okay.
[17:18.250 --> 17:20.210]  A couple more color spaces.
[17:20.450 --> 17:22.430]  This is one done by Edward Munzel.
[17:22.570 --> 17:27.030]  And this was done, I think, in the first part of the last century.
[17:27.370 --> 17:29.250]  I think 1930s and 40s.
[17:29.330 --> 17:38.230]  And he actually asked people to judge different colors by giving them different color swatches and saying, hey, can you tell the difference of this?
[17:38.370 --> 17:41.090]  And so this is a perceptually based color space.
[17:41.210 --> 17:43.290]  And you'll notice that it's not a full circle.
[17:43.290 --> 17:46.570]  It's the same kind of hue and lightness towards the middle.
[17:46.850 --> 17:49.690]  And saturation increases towards the outside or brightness.
[17:49.930 --> 17:52.330]  And it's not fully populated.
[17:52.570 --> 17:55.570]  And this is basically for the same reason...
[17:55.570 --> 17:56.750]  It keeps jumping ahead.
[17:57.970 --> 18:00.330]  That, you know, yellow looks brighter than blue.
[18:00.330 --> 18:01.150]  Just perceptually.
[18:01.210 --> 18:02.070]  Just the way it does.
[18:02.290 --> 18:03.050]  So you can get these...
[18:03.050 --> 18:08.310]  You get kind of these bright yellow on the brightness axis, which is away from the center.
[18:08.310 --> 18:10.190]  And a little less of the blues.
[18:10.930 --> 18:11.430]  All right.
[18:11.570 --> 18:13.650]  So here's a nice gallery of color pickers.
[18:13.790 --> 18:18.870]  If you've used Photoshop or anything, you're probably familiar with these.
[18:19.030 --> 18:24.050]  And most of these give you ways to navigate things in the hue, saturation, and lightness thing.
[18:24.910 --> 18:25.510]  All right.
[18:26.230 --> 18:29.250]  Well, I'm going to talk about this color picker.
[18:29.250 --> 18:31.370]  And this is kind of really cool.
[18:31.570 --> 18:34.890]  And it took me a while to understand exactly what was going on.
[18:35.010 --> 18:36.670]  So I thought I'd share it with you because it's pretty cool.
[18:37.030 --> 18:40.530]  Has anybody seen this thing before just coming across?
[18:40.950 --> 18:41.270]  Right.
[18:41.450 --> 18:44.050]  So it's like, what's this weird kind of egg-shaped thing?
[18:44.050 --> 18:45.950]  And it's got this...
[18:47.570 --> 18:52.230]  On the outside of this, at least one edge of it, it has these numbers.
[18:52.910 --> 18:56.250]  And these, once again, these are the wavelengths of the various colors or something.
[18:56.390 --> 18:58.230]  So you have red here, deep red.
[18:59.210 --> 19:00.370]  Here's yellow, green.
[19:00.690 --> 19:04.110]  The colors here are only really approximate inside this egg shape.
[19:04.290 --> 19:08.290]  And down here to blue, 480 and 450.
[19:08.290 --> 19:08.830]  Okay.
[19:09.070 --> 19:16.750]  So the way this works, this is kind of a recipe for making your own colors.
[19:16.930 --> 19:17.470]  If you want...
[19:17.470 --> 19:25.730]  If you have particular sources of colors and you mix them together, this particular chart will tell you what kind of colors you're going to wind up with.
[19:25.930 --> 19:41.690]  So, for instance, if you have a red pixel and a blue pixel, if you have some kind of, say, damaged monitor or something, and you can only show red and blue pixels, then it turns out that using this chart, you can tell what kind of colors you can get out.
[19:41.850 --> 19:49.590]  And the kinds of colors you can get out are everything that's on the line between your two color sources, in which case here it's red and blue.
[19:50.150 --> 20:05.110]  And furthermore, it's kind of proportional in that if you have basically 70% blue and 30% red, then you will wind up at this point here, which is kind of this bluish purple.
[20:05.110 --> 20:16.790]  And this, once again, kind of illustrates how you can mix red and blue to get this non-spectral color, magenta, which doesn't have a corresponding thing.
[20:19.570 --> 20:21.050]  And...so this is neat.
[20:21.510 --> 20:36.770]  And this...people who do research in color vision, people who build monitors, people like the Society of Motion Picture and Television Engineers who really care about this stuff, have this thing called a gamut.
[20:36.970 --> 20:53.510]  And this is basically if you have a monitor, if you have a camera, if you have sensors, if you have things that detect kind of colors, if you have a monitor and the red pixel is this particular color, and the green pixel is this particular green, and the blue pixel is this particular blue,
[20:53.770 --> 20:59.250]  then you can show any color that's inside this triangle, but you can't show any color that's outside.
[20:59.710 --> 21:01.530]  Boy, this keeps advancing.
[21:01.890 --> 21:02.710]  Apologize for that.
[21:03.030 --> 21:05.150]  So, this is called the gamut.
[21:05.590 --> 21:11.870]  And this really characterizes a display device in terms of the kinds of colors you can actually get out of it.
[21:12.670 --> 21:14.490]  And this is important for a lot of people.
[21:14.610 --> 21:19.470]  And it's really interesting to know that there's this huge gamut that most things can't display.
[21:19.690 --> 21:27.470]  There's really saturated colors, really saturated greens and blues and reds that you're just not going to get out of a projector.
[21:28.630 --> 21:41.850]  There are some now, some displays which have a fourth color, like an amber thing, which instead of a triangle give you the quadrilateral, which gives you more possible colors that you can get at.
[21:43.670 --> 21:49.530]  Okay, so let's talk a little bit about how they came up with this weird egg-shaped thing.
[21:50.290 --> 21:53.150]  And I apologize, I'm going to throw a little bit of math at you.
[21:53.210 --> 21:54.050]  I know it's kind of early.
[21:54.810 --> 21:57.030]  Okay, so how did they come up with this thing?
[21:57.110 --> 22:00.010]  Well, basically, this was done back in the 30s.
[22:00.090 --> 22:03.210]  And they did a pretty careful scientific experiment.
[22:03.410 --> 22:08.390]  They would show people a fixed wavelength, a pure color, that they got from someplace.
[22:08.390 --> 22:19.910]  And they would give them basically three knobs to adjust these three colors to mix in the red, green, and blue and get a particular value for any of them.
[22:19.990 --> 22:25.490]  For example, if you saw orange, you would probably mix a lot of red and a little bit of green and no blue.
[22:25.690 --> 22:29.390]  And then you would get a value for red, green, and blue, which corresponds to that orange.
[22:31.450 --> 22:39.690]  Okay, so then what they did is they take this, and you basically got these tri-stimulus values, as they're called.
[22:39.930 --> 22:41.870]  And there's some weird stuff here, like this is negative.
[22:42.130 --> 22:43.850]  How do you get a negative light?
[22:43.990 --> 22:49.150]  Well, it turns out what you do is you just add a little bit to everything else, and so you can take down the red.
[22:49.290 --> 22:54.710]  So you add an offset to the blue and the green, and that gives you the equivalent of a negative red.
[22:54.710 --> 22:59.890]  And you can basically divide these things, and you get these two values here.
[23:00.150 --> 23:06.330]  This is a two-dimensional plot, and the way it works is basically, the X is basically how much red you have.
[23:06.510 --> 23:10.290]  X is red versus blue, and Y is green versus blue.
[23:10.390 --> 23:13.450]  If both X and Y are small, you have blue.
[23:13.610 --> 23:15.650]  If X is large, you have red.
[23:15.950 --> 23:17.570]  If Y is large, you have green.
[23:17.570 --> 23:26.990]  So this is kind of... the other color spaces, we were also concerned with kind of brightness and darkness, saturation and lightness.
[23:27.310 --> 23:30.030]  This, we're just taking a slice through the color space.
[23:30.170 --> 23:32.090]  We're not worrying about how bright things are.
[23:32.190 --> 23:34.330]  We're assuming that that's given.
[23:35.850 --> 23:36.270]  Okay.
[23:36.270 --> 23:39.690]  So, there's more good stuff in the CIE chart.
[23:39.810 --> 23:46.490]  And if you look right in the center, the center is where the white stuff is, basically, where you mix all the colors together in the center, you get white.
[23:46.510 --> 23:50.670]  Or if you mix kind of colors that are opposite colors, you can also get white.
[23:50.670 --> 23:59.170]  And in the center of that is where all the... what's called the black-body locus.
[23:59.310 --> 24:00.830]  And this is the black-body radiation.
[24:01.050 --> 24:04.650]  If you take a piece of iron and you heat it up with a blowtorch, it's gonna glow red.
[24:04.750 --> 24:08.150]  And you heat it up hotter with an oxyacetylene torch, it's gonna glow white.
[24:08.430 --> 24:11.610]  Well, that's called the black-body locus.
[24:11.770 --> 24:14.490]  And that's kind of on this curve here, in the center here.
[24:14.610 --> 24:16.270]  And you can see these temperatures here.
[24:16.650 --> 24:18.070]  In Kelvin, they're pretty hot.
[24:18.070 --> 24:24.290]  But those are, in fact, the temperatures of the colors that you would get where you do heat an ideal black-body to that particular temperature.
[24:24.970 --> 24:30.850]  And there's... if you have a fancy monitor, you know you can play with the color temperature of your monitor.
[24:30.970 --> 24:36.750]  And that just basically makes it a little bluer, a little redder, just by shifting the colors back and forth along this location.
[24:37.890 --> 24:44.030]  And if you're a photographer, there's something called D50, which is kind of the preferred white.
[24:44.030 --> 24:51.590]  It's the... it's the standard white that you use to calibrate photographic equipment.
[24:51.790 --> 25:01.450]  And if you're an astronomer, stars are pretty good black-bodies, and they... the colors of stars, depending on their temperature, also lines up on this... on this curve here.
[25:01.670 --> 25:11.730]  And there's things called daylight spectrum, and then there's incandescent spectrum, which is a little redder, a little warmer, a little less blue in it.
[25:12.950 --> 25:22.650]  Okay, so... this is the... this is kind of the whole enchilada with the gamut, and the colors, and the spectrum, and the black-body radiation.
[25:22.890 --> 25:28.930]  So this is... this is interesting, really, to me, as kind of a map of the color perception system.
[25:30.230 --> 25:33.050]  All right, so... how do you make colored light?
[25:33.170 --> 25:37.670]  Well, I can't go into all of these just because there's a million different ways, but lots of them are super interesting.
[25:38.170 --> 25:39.370]  You can do filters.
[25:39.650 --> 25:42.890]  You can do dichroic filters.
[25:43.190 --> 25:47.770]  A lot of stage lights use dichroic filters, which is an interference effect.
[25:47.770 --> 25:48.910]  It's not actually a pigment.
[25:49.370 --> 25:52.930]  There's optical refraction, just like the prism.
[25:53.110 --> 25:54.630]  You shine light into it.
[25:54.830 --> 26:02.550]  If your... if your glasses are sufficiently thick, like mine are, if you look out the corner of it, you get what's called chromatic aberration.
[26:02.550 --> 26:05.550]  You can use that as a spectroscope, because you basically get a little prism.
[26:06.670 --> 26:07.830]  Gas discharge tubes.
[26:07.970 --> 26:09.190]  I'll talk about those in a moment.
[26:10.050 --> 26:11.390]  Fluorescent dyes and phosphors.
[26:12.530 --> 26:14.910]  Different pigments, which light up when you zap them.
[26:15.730 --> 26:16.130]  LEDs.
[26:16.670 --> 26:18.270]  If you like building stuff.
[26:18.490 --> 26:19.670]  If you've done stuff with Arduinos.
[26:20.450 --> 26:23.350]  There's a beautiful spectrum of LEDs you can get now.
[26:24.270 --> 26:26.370]  And pigments, and reflection, and inks.
[26:26.450 --> 26:27.590]  Did that make it on the bottom?
[26:27.590 --> 26:39.630]  If you're an artist, if you're into this, there's a wonderful site called handprint.org, which basically talks about the color characteristics of all the incredible pigments that are out there for artists.
[26:39.710 --> 26:40.710]  You know, that have been used.
[26:41.370 --> 26:48.670]  Like the ground up stone, which make this a beautiful blue, to just about anything else.
[26:48.830 --> 26:51.030]  Don't have time to go into that, fortunately, for you.
[26:52.210 --> 27:08.030]  Okay, so, something you don't always notice, but is always the case, is that the light that you use to see things with, also has kind of a color, a color associated with it.
[27:08.190 --> 27:14.150]  And ideally, white light is white, and has all the colors and more or less equally balanced.
[27:14.350 --> 27:16.290]  But that's, quite often, that's not the case.
[27:17.170 --> 27:28.350]  On the left here, this is the spectrum of an incandescent bulb that is not very hot, so you can see it's got a lot of stuff in the red and not much in the blue.
[27:28.910 --> 27:36.270]  This one's supposed to be a light that's closer to sunlight, and sunlight has a peak in the green, and it falls off a little in the blue and the red.
[27:36.270 --> 27:40.110]  Once again, these are the wavelengths of the light that you're seeing here.
[27:44.410 --> 27:52.690]  And a very common source of light is fluorescent bulbs, and there aren't any in here.
[27:52.950 --> 27:57.850]  But in the center of a fluorescent tube is basically a mercury vapor discharge.
[27:58.130 --> 28:00.810]  And this is like a neon tube, only it's got mercury vapor in it.
[28:00.810 --> 28:06.870]  You put a high voltage across it, you get electrical discharge, and this is the spectrum of it.
[28:06.970 --> 28:07.970]  You get these lines.
[28:08.110 --> 28:11.850]  And this is beautiful, pure quantum physics at work, basically.
[28:11.990 --> 28:15.170]  These lines come from energetic transitions.
[28:15.350 --> 28:19.890]  You basically, you get an atom, you hit it with an electron from the discharge.
[28:20.230 --> 28:27.390]  Some of the atoms in the, some of the electrons in that atom go up to a higher energy level.
[28:27.390 --> 28:40.350]  They stay there for a while, and when they fall back down, they release a wavelength of light, and the wavelength of light, in other words, the color of the light that they release, is proportional to the energy distance between that, between those energy orbitals,
[28:40.490 --> 28:42.450]  which is really a function of the element.
[28:43.070 --> 29:02.950]  And so, this is a neat way of not only to figure out what gases are in a particular tube that you happen to be zapping with electricity, but you can figure out the chemical composition of things like stars that are thousands of light years away just by looking at,
[29:02.970 --> 29:06.890]  basically, either absorption or emission lines in the spectrum.
[29:09.570 --> 29:12.350]  So, different gases have different emission lines.
[29:12.450 --> 29:26.590]  This is mercury vapor, and this has blue, strong blue, very strong blue, very strong green, very strong ultraviolet one, you can't see that one, but that's why they use it in fluorescent tubes, because fluorescent tubes have the mercury vapor, and they have a phosphor.
[29:26.790 --> 29:30.870]  The phosphor absorbs the UV, and it re-radiates out as visible light.
[29:31.390 --> 29:32.150]  Here's neon.
[29:32.510 --> 29:36.510]  Neon has lots of red, a little bit of green, not much blue.
[29:36.950 --> 29:38.650]  And you're probably familiar with this.
[29:39.790 --> 29:48.210]  For a while, they were using mercury vapor lamps for street lights, because they're pretty efficient, they give you a lot of light, but they don't have a lot of red in them.
[29:48.570 --> 30:00.130]  So, if you were out under a mercury vapor street lamp, and there was a red car, it would be this weird brownish color, which your brain would still make kind of red, even though you were seeing actually very little red.
[30:00.890 --> 30:03.470]  Nowadays, they use low-pressure sodium.
[30:03.870 --> 30:05.070]  It's kind of the same thing.
[30:05.170 --> 30:06.890]  Low-pressure sodium is very little blue.
[30:07.110 --> 30:16.550]  So, if you see a blue car, next time you're under a street lamp, look at a blue car and think about what color is that that you're actually seeing, because there's not a lot of blue in it.
[30:18.370 --> 30:25.910]  Okay, so this is a little thing I did just to contrast a mercury vapor discharge with some neon things.
[30:26.050 --> 30:28.150]  They're really kind of interesting complementary colors.
[30:30.450 --> 30:30.890]  Okay.
[30:30.890 --> 30:32.990]  So, here's a little bit of news you can use.
[30:32.990 --> 30:45.070]  You can measure how good lights are to basically give you the colors that you want to see by this thing called the CRI, or the color rendering index.
[30:45.710 --> 30:52.110]  And if you go to the hardware store and you buy a light bulb now, you buy one of those compact fluorescents, that will typically have this number on it.
[30:52.230 --> 30:53.830]  And it goes from 0 to 100.
[30:54.190 --> 30:56.690]  100 is perfect color rendering index.
[30:56.850 --> 30:58.690]  That means it's just like sunlight.
[30:58.690 --> 31:05.950]  All the colors that you can see under this lamp are going to be nice and saturated and things like that.
[31:06.190 --> 31:10.190]  And different kinds of illumination have different color rendering indexes.
[31:10.310 --> 31:18.510]  Like this I stole from a website that's trying to sell you LEDs, and they say, oh, look at the CRI is much better than fluorescent lighting.
[31:18.670 --> 31:28.290]  Now, fluorescent lighting, basically you have the blue from the mercury vapor discharge, and you've got this orange-yellow from the phosphor.
[31:28.390 --> 31:34.690]  There's not a lot of red in fluorescent lighting, which means reds look kind of washed out and a little bit of weak.
[31:34.850 --> 31:47.130]  And I've actually noticed a lot of supermarkets are using high-pressure gas discharge things, because they're efficient and they have a really good color rendering in the reds.
[31:47.150 --> 31:50.010]  You go into the produce, it's got lots of red, lots of green.
[31:50.010 --> 31:52.070]  You go into the produce section and the stuff just pops.
[31:52.210 --> 31:52.850]  It looks beautiful.
[31:54.750 --> 32:03.890]  So there's a lot of kind of psychological things that people use to maybe influence your choice.
[32:03.910 --> 32:07.050]  So if you're aware of this, this is good stuff to know.
[32:07.490 --> 32:10.230]  Okay, so how do you figure out this color rendering index?
[32:10.490 --> 32:11.370]  You know, it's kind of hard.
[32:11.590 --> 32:19.310]  Well, it's kind of hard just to look at this and say, well, yeah, well, the reds look washed out, but you can actually characterize it.
[32:19.690 --> 32:33.910]  And the way you do this is you take the reference, you take a bunch of colors, and I think there's 16 standard colors which they use, and you illuminate the reference colors with a known good white light source, like a nice halogen which has good color rendering.
[32:34.170 --> 32:40.370]  And then you illuminate the same colors with the source of light you're testing.
[32:40.610 --> 32:46.190]  And then you can figure out the colors that you see on the CIE chart, right?
[32:46.190 --> 32:48.550]  Those will wind up in a particular place.
[32:48.790 --> 32:51.370]  And so you can tell the difference that they've moved.
[32:51.510 --> 32:54.430]  Hopefully, if you have good color rendering, they won't move at all.
[32:54.830 --> 32:57.370]  If you have bad color rendering, they will move a little bit.
[32:57.490 --> 33:10.870]  And so the actual number you get is basically you take the average of the distance that they've moved in this CIE color space for these 16 colors, multiply it by 4.6 and subtract it from 100.
[33:10.870 --> 33:15.550]  So that gives you a number theoretically between 0 and 100.
[33:15.550 --> 33:18.030]  Although if it's really bad, I think you can get negative ones.
[33:19.390 --> 33:19.870]  Okay.
[33:20.490 --> 33:32.610]  So here's something that you may have noticed if you bought, if you ever bought like one of those cheap LED flashlights, the color rendering from those is not very great.
[33:32.730 --> 33:38.430]  And the reason that that happens is basically there's two sources of light in a white LED.
[33:38.430 --> 33:46.250]  There's a very bright blue LED and then there's this yellow greenish phosphor, which the blue LED excites.
[33:46.390 --> 33:47.970]  So you get this yellow green plus blue.
[33:48.090 --> 33:49.250]  There's not a lot of red in it.
[33:49.810 --> 33:52.990]  And you can see what's going on here in the chromaticity chart.
[33:53.010 --> 33:57.530]  If I take yellow green and I mix it with blue, I do get this stuff that's white.
[33:58.310 --> 34:04.330]  And it's a pretty white white, but the color rendering is pretty lame because there's not much red in it.
[34:04.390 --> 34:07.490]  So that explains why stuff looks weird under LEDs.
[34:07.490 --> 34:12.030]  And there's a lot of research into new phosphors that have more red in them.
[34:12.530 --> 34:15.750]  The same thing for compact fluorescent bulbs.
[34:15.990 --> 34:18.250]  A lot of the colors on those look a little weird.
[34:18.390 --> 34:20.550]  And that's basically just because they don't have good phosphors.
[34:21.630 --> 34:22.230]  All right.
[34:23.050 --> 34:23.530]  Okay.
[34:23.650 --> 34:25.890]  So let's talk a little bit more about perception.
[34:26.230 --> 34:26.410]  All right.
[34:28.390 --> 34:30.430]  Adelson, he's a professor at MIT.
[34:30.430 --> 34:34.350]  He has this wonderful illusion, and you've probably seen this.
[34:34.470 --> 34:37.570]  If you haven't, look at the squares marked A and B, right?
[34:37.790 --> 34:40.530]  And clearly A looks darker than B, right?
[34:41.010 --> 34:44.570]  Well, the secret to this is A is exactly the same color as B.
[34:44.770 --> 34:48.850]  And that seems kind of counterintuitive, but there's the proof right there.
[34:48.850 --> 34:51.970]  And here I've even taken away.
[34:52.110 --> 34:56.020]  Those are exactly perfectly half level gray scales.
[34:56.730 --> 35:05.950]  So even when your eye is detecting things correctly, there's a whole lot of higher level processing going on to make sense of the world.
[35:05.950 --> 35:09.630]  And the same thing happens with color.
[35:10.170 --> 35:11.830]  This is a little demonstration.
[35:12.010 --> 35:13.810]  This is actually an interactive thing on my website.
[35:15.410 --> 35:20.030]  These two, you can adjust the colors of the surrounding things.
[35:20.110 --> 35:22.230]  You can adjust the colors of the internal boxes.
[35:22.470 --> 35:25.290]  And I've kind of adjusted these so they kind of look the same.
[35:25.670 --> 35:28.290]  I don't know how it does on the projector.
[35:28.450 --> 35:29.890]  But it turns out they're not the same.
[35:30.110 --> 35:35.550]  So the surrounding colors can really influence your perception of things.
[35:36.490 --> 35:37.090]  All right.
[35:38.650 --> 35:39.090]  Okay.
[35:39.230 --> 35:41.330]  One more little kind of neat thing.
[35:41.550 --> 35:44.930]  I hope everybody has played with laser pointers.
[35:45.510 --> 35:49.050]  And they've seen this thing called laser speckle, which is this...
[35:49.050 --> 35:52.870]  I apologize, I don't have my green laser otherwise I can show it to you right here.
[35:53.090 --> 35:58.530]  It's got this really interesting graininess effect that you see from laser light.
[35:58.690 --> 36:00.610]  And that happens because laser light is monochromatic.
[36:01.310 --> 36:05.470]  It's got basically one wavelength and there's not a lot of bandwidth.
[36:05.550 --> 36:07.170]  So you get one pure color.
[36:08.090 --> 36:10.190]  And not only that, it's all lined up.
[36:11.030 --> 36:15.890]  And it's in phase, which means the peaks line up with the peaks and the troughs line up with the troughs.
[36:16.210 --> 36:29.750]  So what happens with the laser speckle is when you shine it on an irregular surface, and this is irregular on the optical wavelength scale, which is pretty much anything, you get constructive and destructive interference.
[36:29.750 --> 36:37.690]  So sometimes the things will add up and the waves will be on top of the waves and sometimes you'll add them up and the waves will be on top of the troughs and they'll cancel out.
[36:37.790 --> 36:39.730]  So you'll get the light and the dark things.
[36:39.870 --> 36:42.430]  And if you ever see this thing, a couple things to notice.
[36:42.610 --> 36:44.850]  If you wear glasses, take off your glasses.
[36:45.150 --> 36:46.090]  It's really interesting.
[36:46.230 --> 36:54.210]  You'll still see the speckle no matter how bad your eyes are because the speckle is happening happening on your retina, not on the object itself.
[36:55.450 --> 37:06.370]  And another thing to notice, even if your eyesight is good, if you move your head left, the appearance will move right and vice versa and that's because it's happening again on your retina.
[37:06.990 --> 37:10.190]  Alright, I'm running out of time.
[37:10.330 --> 37:12.630]  This is a last illusion here to the Oka.
[37:12.830 --> 37:17.070]  I apologize once again if you had an entertaining evening last night.
[37:17.350 --> 37:24.410]  I wish I had a good explanation for this but there's all sorts of things you've seen as far as equiluminant colors.
[37:25.210 --> 37:30.430]  This one obviously has some frequency effects, high spatial frequency, low spatial frequency.
[37:31.570 --> 37:35.770]  If you Google Kita Oka, he's got a page on his site.
[37:35.850 --> 37:37.330]  I think he's a professor in Japan.
[37:37.590 --> 37:38.870]  He's got a lot of these things.
[37:39.010 --> 37:39.870]  They're kind of amazing.
[37:41.170 --> 37:43.370]  And that's pretty much my talk.
[37:43.550 --> 37:45.010]  I have some time for questions.
[37:52.720 --> 37:55.460]  Okay, I can't really see so if you've got a question.
[37:55.560 --> 37:55.840]  Yes, sir?
[37:55.840 --> 38:09.900]  When you're talking about the color gamut and how much saturation is on your sensor projector, is there any software or like menus that you know of or like build cameras that you can see or range on your sensor, like how much saturation you're getting?
[38:10.580 --> 38:12.660]  Yeah, there's... Oh, sure.
[38:13.460 --> 38:14.060]  Thank you.
[38:14.220 --> 38:14.300]  Yeah.
[38:14.440 --> 38:19.320]  The question was, how can you tell how good the colors are on your monitor, how saturated they are?
[38:20.020 --> 38:30.640]  Yeah, I think there's... I'm not an expert at this, but there's a whole industry for graphics designer making sure what you see on your computer screen is the same stuff that gets printed.
[38:30.880 --> 38:34.400]  And there's something called Color Monkey, I think, M-O-N-K-E-E.
[38:34.560 --> 38:35.940]  Does anybody heard of this?
[38:36.380 --> 38:38.900]  It gives you a little device you can hold up to your screen.
[38:39.000 --> 38:40.300]  It's got a color sensor on it.
[38:40.540 --> 38:47.420]  It does some software that runs through the colors and it will actually characterize your particular display for how good it is at rendering colors.
[38:47.580 --> 38:56.780]  And then you can set a color profile so basically the colors that you see on your screen are the same as the colors that you get from the shop where you send your stuff to get printed.
[38:59.480 --> 39:00.120]  Other question?
[39:00.360 --> 39:00.480]  Yes, sir.
[39:00.500 --> 39:04.360]  Do you know any of the reason why we perceive red is red, blue is blue, green is green?
[39:04.600 --> 39:07.320]  Why do we perceive red is red, blue is blue, green is green?
[39:07.500 --> 39:08.240]  I don't.
[39:08.360 --> 39:11.120]  And that's a really interesting philosophical discussion.
[39:14.060 --> 39:28.660]  On a similar comparison of the perception of color and turning light into printed, why is it that you mix red and yellow to make orange?
[39:40.820 --> 39:41.740]  That's right.
[39:41.900 --> 39:47.340]  So the question is why is light additive and why are pigments subtractive?
[39:48.080 --> 39:49.340]  Or at least different.
[39:49.400 --> 39:49.620]  Right.
[39:49.800 --> 39:50.340]  Why are they different?
[39:50.620 --> 39:50.720]  Okay.
[39:55.720 --> 39:57.520]  Yeah, that's actually not the...
[39:57.520 --> 39:59.080]  Was that me?
[40:03.640 --> 40:04.420]  Fire alarm?
[40:05.520 --> 40:07.480]  Someone hacked the fire alarm.
[40:07.480 --> 40:07.540]  Someone hacked the fire alarm.
[40:08.080 --> 40:08.320]  All right.
[40:09.060 --> 40:09.540]  Okay.
[40:09.600 --> 40:14.660]  So I'm going to try and talk through this until there is actual demonstration of actual emergency.
[40:15.300 --> 40:16.340]  Just go back here.
[40:22.870 --> 40:23.390]  Right.
[40:25.930 --> 40:38.050]  Okay, so I obviously did a lame job of explaining this, but subtractive colors basically, what they do is they remove wavelengths of light.
[40:38.290 --> 40:41.510]  So a yellow color is actually removing the blue.
[40:41.950 --> 40:50.010]  So the reason it looks yellow is you're shining white on it, it's taking out the blue and it's leaving yellow, which is a mix of red and green.
[40:52.470 --> 40:54.150]  Did that help at all?
[40:55.490 --> 40:56.010]  Okay.
[40:56.530 --> 40:57.910]  The same thing with cyan.
[40:58.070 --> 41:04.070]  Cyan removes red, leaving the blue and the green, which is cyan.
[41:04.230 --> 41:08.370]  So this is why when printing you use these colors which are subtractive, which remove things.
[41:08.950 --> 41:17.330]  If you mix red and green pigments, if you mix them all together, you get this muddy brown color.
[41:21.190 --> 41:24.810]  I've actually read about people who have a fourth cone.
[41:25.030 --> 41:25.950]  Yes, yes.
[41:26.170 --> 41:30.790]  The question is, he's heard about people who have a fourth cone.
[41:31.030 --> 41:39.710]  These people are called tetrachromats and so they can see theoretically further into the visible spectrum.
[41:39.710 --> 41:41.050]  I've heard of that too.
[41:41.370 --> 41:42.810]  I don't know much about it.
[41:42.910 --> 41:43.570]  I don't...
[41:44.390 --> 41:53.470]  Apparently, these people tend to be women just for similar genetic reasons like most red, green color blindness tend to be men because that's on the Y chromosome.
[41:53.970 --> 41:56.170]  I don't know much about it but it's super cool.
[41:56.170 --> 42:08.510]  I think the extra pigment is in the yellow region so they can see more saturated colors and different pigments than we would.
[42:10.130 --> 42:10.810]  Yes, sir.
[42:11.010 --> 42:12.470]  Any notes on color blindness?
[42:12.710 --> 42:14.230]  Any notes on color blindness?
[42:14.390 --> 42:25.130]  Yeah, I didn't get into that but a simple explanation for red, green color blindness is that one of your sets of cone cells doesn't work.
[42:25.290 --> 42:25.850]  Right?
[42:26.090 --> 42:34.190]  So the distinction between red and green is kind of lost and actually really common specifically among men.
[42:34.410 --> 42:39.750]  There are other types of color blindness and things like that.
[42:40.530 --> 42:41.850]  That's always damage.
[42:42.610 --> 42:52.470]  To my knowledge, once again, I'm not a biologist, not my area of expertise but the most prominent case is that.
[42:53.770 --> 42:54.450]  Two things.
[42:54.690 --> 42:54.870]  Yes.
[42:55.110 --> 42:57.530]  First of all, you don't need to evaluate that was a false alarm so we're all close.
[42:57.570 --> 42:57.790]  Okay.
[42:58.870 --> 43:05.390]  Second of all, I was really interested in the stuff about the burning colors.
[43:05.990 --> 43:18.450]  I forget like the exact term but can you talk about how that relates to visible light at all because I sort of didn't catch the black mass capture.
[43:19.030 --> 43:20.230]  Oh, right, right, right.
[43:20.530 --> 43:20.850]  Okay.
[43:21.230 --> 43:26.510]  The black body spectrum, this is the light you get when you heat something.
[43:27.410 --> 43:33.790]  And the scientists like to use a black body because that doesn't have any, that's a perfect radiator.
[43:34.090 --> 43:35.150]  But you can do the same thing.
[43:35.270 --> 43:41.610]  If you take a piece of iron and you heat it with a torch and it actually, let me see if I can find that.
[43:46.290 --> 43:49.790]  Actually, you can do this yourself at home with a light bulb and a light dimmer.
[43:50.890 --> 43:53.850]  If you, yeah.
[43:54.030 --> 44:04.670]  So if you take a light bulb and you put it on a super low setting so it's just kind of orange-ish and things, then you'll be emitting light with a color temperature that's down here.
[44:04.670 --> 44:11.810]  And as you increase the power to the bulb, it'll get less and less red and more and more white until some, some place you're around daylight here.
[44:13.630 --> 44:16.050]  And so here's an example of this.
[44:16.190 --> 44:17.830]  This is a, this is a light bulb.
[44:18.110 --> 44:21.470]  And now notice that it is giving you a full spectrum.
[44:21.650 --> 44:23.990]  It's giving you colors of pretty much all wavelengths.
[44:24.230 --> 44:29.070]  But a lot of the wavelengths are in the red spectrum.
[44:29.070 --> 44:31.510]  And with light bulbs, light bulbs are really inefficient.
[44:31.910 --> 44:34.050]  Most of their output is actually in the infrared.
[44:34.050 --> 44:38.230]  So you're wasting all that energy on infrared light that you can't even see.
[44:38.430 --> 44:40.110]  So, which is why they're warm to the touch.
[44:40.810 --> 44:45.390]  So, so if you go back to your, I think, one day before with the temperatures on it.
[44:45.990 --> 44:48.470]  That, well that, or the next one.
[44:48.630 --> 44:51.170]  So that curve would be different for different bodies.
[44:52.330 --> 44:54.090]  Can I have your attention?
[44:54.510 --> 44:55.870]  Can I have your attention please?
[44:56.470 --> 44:57.510]  We are just
[45:05.700 --> 45:07.900]  a test on the fire alarm system.
[45:08.320 --> 45:11.680]  Please disregard any signals that you may hear or see.
[45:12.260 --> 45:15.510]  We have to test.
[45:15.790 --> 45:17.150]  Pay no attention to the little man.
[45:17.430 --> 45:17.490]  Right.
[45:18.030 --> 45:18.510]  Okay.
[45:18.910 --> 45:25.610]  So, yeah, well the, the, the physics behind this is that basically anything you heat up this hot will give you color that temperature.
[45:25.750 --> 45:35.210]  The physicists like to use this ideal black body thing, which has this perfect property, which it adheres to the equations exactly.
[45:35.390 --> 45:40.550]  Actually, this is, this is Max Planck who solved the ultraviolet catastrophe by figuring all this out.
[45:40.930 --> 45:42.810]  This is awesome quantum physics here.
[45:43.410 --> 45:46.990]  But this is, I'm sorry, did I answer the question?
[45:47.130 --> 45:52.230]  Well, I was just, I was just thinking that you heat iron to a glowing red.
[45:52.550 --> 45:52.770]  Yeah.
[45:52.770 --> 45:53.710]  The color of red.
[45:53.850 --> 45:53.990]  Right.
[45:53.990 --> 45:55.430]  It's gonna be at that temperature.
[45:55.570 --> 45:55.870]  Exactly.
[45:56.050 --> 45:57.310]  But you heat titanium.
[45:57.590 --> 45:57.910]  Right.
[45:59.630 --> 46:01.070]  Yeah, there are slight differences.
[46:01.290 --> 46:04.970]  I think in practice, in practice, it's, it's, it's not large.
[46:05.070 --> 46:08.170]  Only, only if you're doing like real super science do you really care.
[46:08.330 --> 46:08.510]  Okay.
[46:08.690 --> 46:10.190]  Most things are pretty good black bodies.
[46:10.970 --> 46:11.130]  Question.
[46:16.070 --> 46:17.270]  Yes, sir.
[46:26.120 --> 46:26.720]  Right.
[46:27.780 --> 46:30.200]  Color detection methods.
[46:30.780 --> 46:34.960]  Well, there's, I did talk about the three types of cones.
[46:35.080 --> 46:37.000]  And that's, that's, that's the one to your eye.
[46:37.700 --> 46:45.480]  Any light sensor, anything which will detect light, you can use as a color sensor by putting a color filter in front of it.
[46:47.720 --> 46:57.400]  So, for example, this, this color monkey thing that you can put up against your screen and detect the colors it's actually putting out.
[46:57.640 --> 46:59.480]  I'm guessing, I haven't taken one of these apart.
[46:59.620 --> 47:04.720]  I'm guessing that has three photo transistors or photo diodes with three filters on it.
[47:04.720 --> 47:07.500]  Probably a, a red, green, and blue filter.
[47:08.120 --> 47:10.560]  Or just like the, the sensors in your camera.
[47:10.720 --> 47:17.220]  They're CCD or CMOS and they have little filters on top of them which filter out the red, green, and blue.
[47:17.360 --> 47:20.500]  In fact, I think that was that slide that I didn't get to.
[47:22.620 --> 47:24.040]  Let's see if I can find that again.
[47:24.680 --> 47:25.400]  From the,
[47:29.920 --> 47:44.020]  yeah, here's a, here's a, here's a response from a particular Kodak camera which shows you basically, um, uh, the response to, uh, cyan light, uh, magenta light, yellow light, and cyan once again is just the sum of the green and the blue.
[47:44.260 --> 47:52.400]  So, if you look at the green, which is this green thing, and the blue thing which is the dots, you sum them together, you get the cyan thing.
[47:52.820 --> 47:53.940]  Did that answer the question?
[47:54.100 --> 47:54.640]  Well, yeah.
[48:14.560 --> 48:15.000]  Right.
[48:15.100 --> 48:27.980]  There's also a scientific instrument called, um, a spectrograph where you use a diffraction grating, and you, um, put whatever light you're looking at it, and you, you vary the diffraction grating, which basically lets you sweep across the spectrum, then you can measure the,
[48:28.260 --> 48:31.580]  um, you can measure the response at any particular wavelength exactly.
[48:32.640 --> 48:33.140]  That's it.
[48:33.420 --> 48:33.740]  Question?
[48:33.920 --> 48:37.540]  What does that do for the, um, for the, um, non-spectral colors?
[48:38.360 --> 48:41.140]  When you remove one of these, some of these non-diffraction grating?
[48:41.140 --> 48:42.320]  Do you have a website or something?
[48:43.260 --> 48:44.360]  Uh, yes I do.
[48:44.520 --> 48:48.740]  Unfortunately not, most of this isn't on it, so I apologize.
[48:48.740 --> 48:48.900]  There it is.
[48:50.080 --> 48:51.120]  Oh, well thanks.
[48:52.160 --> 48:59.060]  Okay, so, I will, I will, so the question was, what does the spectrophotometer do for the non-spectral colors?
[48:59.480 --> 49:00.560]  Um, basically nothing.
[49:00.980 --> 49:09.020]  Um, it will show you exactly what the spectral response is of your light, but, um, uh, the non-spectral colors is kind of a perceptual phenomenon.
[49:09.300 --> 49:14.840]  So, if you're seeing this nice magenta, your spectrophotometer will show you lots of blue and lots of red.
[49:18.340 --> 49:18.780]  Question?
[49:18.780 --> 49:21.840]  Yeah, so, uh, I know you talked briefly about UV perception.
[49:22.120 --> 49:22.340]  UV, yeah.
[49:22.480 --> 49:27.640]  We actually have, uh, the cones can perceive the UV greater than the, the eyes themselves, the lens, which filters.
[49:27.840 --> 49:31.320]  So the question is, uh, can you talk a little bit more about UV damage to the eyes?
[49:31.420 --> 49:35.400]  Because I know that that, that that's a wavelength that can cause significant damage to the eyes.
[49:35.760 --> 49:36.580]  How does that work?
[49:36.700 --> 49:38.020]  And what's the, what's the physics involved?
[49:38.180 --> 49:38.740]  What's the biology?
[49:39.140 --> 49:39.500]  Sure.
[49:39.800 --> 49:45.680]  Um, uh, the question was, uh, can you talk about UV, UV response and specifically damage to the eye?
[49:45.680 --> 49:53.140]  Um, well, once again, it's a little bit out of my area of expertise, but, um, uh, UV, um, is strongly, um, active.
[49:53.340 --> 49:56.260]  You know, uh, energy is proportional to H nu, right?
[49:56.360 --> 49:59.400]  That's a, um, that's a, uh, uh, uh, a small wavelength.
[49:59.560 --> 50:01.780]  So that's got lots of energy as you go up in the spectrum.
[50:01.780 --> 50:07.000]  So, um, and the, uh, uh, uh, ultraviolet, uh, spectrum of the sun is damaging.
[50:07.320 --> 50:10.040]  Just go out without sunblock for too long and you will find that out.
[50:10.520 --> 50:15.400]  Um, uh, one of the ways that ultraviolet damages things is by cross-linking.
[50:15.760 --> 50:25.280]  Um, if you've got a polymer like, uh, the, um, the proteins in your lens, um, the polymer's long-changed molecule, it gets damaged by UV by cross-linking.
[50:25.440 --> 50:31.100]  Basically, uh, long chains that are supposed to slide together have these cross-links which kind Kind of bind them and it gets stiff.
[50:31.320 --> 50:36.820]  This is one of the reasons that you lose your ability to accommodate when you get to be a certain age.
[50:37.300 --> 50:41.080]  Another thing is that it will yellow the lens.
[50:41.680 --> 50:48.980]  Noticeably, if you take a piece of plastic and you leave it out for a while, some plastics will yellow.
[50:49.460 --> 50:54.580]  And if there's a chemist or something, a photochemist, they can probably explain that phenomenon a lot better than I can.
[50:56.800 --> 50:58.060]  You notice this everywhere.
[50:58.240 --> 51:06.180]  Like if you go to a travel office and you see the nice poster that's been there for a while, it will tend to fade towards blue, right?
[51:06.360 --> 51:08.560]  The reds will get washed out.
[51:08.680 --> 51:11.620]  And this is because red and cyan absorbs blue.
[51:11.760 --> 51:16.900]  It absorbs the shorter wavelengths, which have more energy, and it basically just destroys the pigment.
[51:17.040 --> 51:21.740]  So it's actually hard to make a red pigment that has good longevity.
[51:22.620 --> 51:23.180]  All right.
[51:23.300 --> 51:25.080]  Maybe time for one more question?
[51:26.500 --> 51:26.980]  Okay.
[51:27.120 --> 51:28.200]  Maybe zero more questions.
[51:28.380 --> 51:28.560]  All right.
[51:28.560 --> 51:29.760]  Thank you for your time.
[51:37.250 --> 51:37.950]  Thank you, John.
[51:39.650 --> 51:40.670]  A couple of announcements.
[51:41.130 --> 51:42.290]  One o'clock today, the key