The Physics and Psychology of Colour - with Andrew Hanson

[Music] please welcome onto the stage Andrew Hansen at the end of this talk you may be convinced that the four colors shown here are identical color is that crazy so let's press on does anyone remember the dress big social media thing about three years ago somebody shared an image of their wedding dress and the world went wild arguing about its colors we'll go into that I bought the dress my daughter was too young to fit into it but there she is standing next to it and I took a photograph I illuminated it brightly from one side and dimly from the other and in one photograph alone we can pull the bits out and we can see that indeed under some conditions it looked white and gold and under under conditions it looked blue and black a conference was held to discuss all the various implications of this the colored group of Great Britain all of their committee were approached by a number of media outlets I was approached by The Telegraph I had a column written in there my mother was delighted my father was disgusted he said you should have been in the guardian before I move on I'd like to tell you who I work for the National Physical Laboratory or NPL is it's a known in short is and the national measurement Institute for the United Kingdom we have about a thousand people all measuring stuff on your behalf when it was set up about 100 years ago by the Royal Society it was actually well they call it the National Physical Laboratory I think they should have called it the National measurement laboratory we do physical measurements there and for 20 years of my life I was engaged measuring color and light and that's the basis of what I'm going to talk to you about today but the building itself well it keeps everything running for us you may not realize it but we rely on measurement every aspect of 21st century living and before relies on knowing how much you've got of things nuts need to fit bolts you need to know that the ingredients the active ingredients in drugs are the right amount and so they're measured really accurately and most importantly when you buy chocolate you need to know you've got the right amount so weights and measures and all sorts of other measurements are really really important to make the world work and that's what we're there for doing it for the UK set up about a hundred years ago in the corner of bushy park and we moved about ten years ago into a nice big new glamourous white building with 388 separate laboratory spaces some of them very large indeed with very odd things going on all measurement though and since the 1910s we've been involved in measuring color and light electricity has only been brought into our houses for about the last 100 years and it was brought in not to run television or iPads or heaters but light so measuring the efficacy of lighting systems has been in there for pretty much a hundred years now we use color every day and there's a photograph of the outside of this building and in just one scene we can see the use of color in branding and it's really important that those brands once they've been sorted out are shared appropriately with the right colors that your webpage has the right colors that mugs and clothing and printed materials all at the same color so branding is really important we need to communicate the colour of these brands appropriately quality control imagine if your car got dented and needed a respray you need to be confident that the color of the respray fits with the original color of your vehicle and finally legal and safety there are lots of yellow lines and white lines out down this road just outside and there are rules about how visible those lights that those those lines need to be so people go out there and measure them all the time you need that for visibility we also use color to manipulate emotion so there's a lovely view graphic up there of all sorts of colors for brands maybe you recognize some of them and according to the people who created this sir this diagram you can associate particular emotional responses with different colors gets your thinking doesn't it so there's a need to measure and specify color let's quickly pick apart those needs you create a brand or an identification system which really really needs colors to be communicated correctly beyond language to other parts of the world who might be creating things on your behalf you also use it in monitoring processes such as creating paints but also I bear if you go into a shop and you're faced with a huge amount of orange juice if one of those containers contain something that looks slightly different in color to the others that's the one you wouldn't buy so in monitoring processes colors use quite a lot now if you're communicating color you need to talk about absolute values you need to be able to recreate exactly a color remotely and that requires an absolute definition but if you're monitoring changes in things maybe you're just interested in difference values or things relative to the target color now these have their different requirements but they'll still require some precision in color measurement now we are humans and we have a visual response to light light has wave properties and we can split light up according to the the wave properties so I've brought along a little demo here which I hope will work can we dim the lights a bit bit James I've got a very nice fancy lightbulb here and I've got a dazzle the guys over there there we go you see that stuff so we have a mixture of white light coming on here and this thing here acts a bit like a CD have you ever held a CD up to the light and see it all the bright colors coming out of it it has the same sort of thing in a raindrop as well and it splits the light up into its various components so from this direction we have the blue of a light-emitting diode shining through a phosphor that generates other parts of the spectrum through the process of fluorescence and then up here we've got the whole thing starting again because that's the way this system works unfortunately so we've got the next lot of blue overlapping but if we could just look at one spectrum remotely just one spectrum then you'd see it going from violet blue indigo well the colors of the rainbow let's see what I've put up next so and our response to that spectrum varies we don't see all colors evenly with the same amount of response at the red end of the spectrum the amount of energy in red light is not enough to excite the photochemistry of our eyes of an in our retina so we become blind to anything beyond a particular wavelength we can't see infrared some animals can at the other end we can't see ultraviolet ultraviolet radiation has the potential to destroy the subtle chemistry and biology of our eyes so we don't want it to come into our eyes and the filter at the front of our eye is the ciliary lens something normally used for organizing light and focusing it also acts as a filter so we cannot see ultraviolet radiation some animals can and somewhere in between at the peak of the green part of the spectrum that's where our eyes are most sensitive to light so as a physics I would say you need less watts of green light energy to excite a sensation of brightness in the eye than you need for the red and for the blue it lots lots more energy to excite the eye if it's red and blue let's look at two objects I'm gonna do the whole lights down again I bought two objects along here can you kill the lights again thank you right I'll see if I can get a slightly brighter spectrum it's looking a bit dim to me there we go that's right oh that's that's a bit let's do that one okay got two objects in my hand here they're actually bits of gaffer tape and if I go from the blue end of the spectrum can you see that that's quite dark but as I go for a little walk out here cyan green part that this top part that's a piece of green tape looks really really bright as I move into the yellow at some point the two will appear equal lightness is there happening for you and then as I go out into the red the green part becomes dark and the red becomes violently light now it's really interesting that people's responses to light are different the cameras seeing this as well might see these slightly differently so you'd all argue about which point the two look equally bright that's quite normal we are all individuals now as a measurement scientist that's great but where are the numbers and more importantly where's my controller here is so we can measure things and I said that light has a wave property we talked about the wave length of light so I've written those along the bottom there the numbers are actually thousand millionths of a meter or nano meters as we call them and we can only see quite a narrow band of wavelengths from 380 in the in the violet to about 780 nanometers in the red so we we popped these two objects and we measured their spectral reflectance how much like they were reflecting as a function of wavelength and there they are can you see two lines now the Green Line shows us that the green object was reflecting a lot of light at the green part of the spectrum but not much at the other extremes the other line which I've made cyan so that you can all see them nice and clearly is only reflecting light at the red end of the spectrum now both that property is something to do with this material it's nothing to do with a light bulb it's nothing to do the way we see it that's a property of the material let's have a look at the three stages of color production we can have many many different types of light source all with different spectra and they will shine on two different things that diffuse and scatter and reflect and absorb and transmit light differently as well again with a spectral property and then we resolve those spectra with our eyes and if we need to we train artificial intelligences or robots or cameras to see color the way that we do so color is a three-stage process something to do with the light source something to do with something that changes the color of the light source by absorbing or whatever light and then the way we see it as well a three-stage process those are quite an odd thing that can can happen this observer is looking at two things and they pay the same okay they appear the same and this observer sees them and they appear different and the reason is that the eyes sense things differently okay can you see that and we call that observer metamerism now let's move to a different situation where you have one eye looking at two things and we change the color of the light source then we have a sample metamerism issue and you might remember when cloth you went into Marks and Spencers or whatever and you compare two things and they would look ident in color under one type of lighting but different in color under a different type of lighting I could see a lot of people nodding their heads oh they'll only match under all the different types of light source if they actually reflects light identically and it's quite unusual to get things to do that so here's your typical ol 'evil question GCSE these days if you have a blue object illuminated with a blue light what color will it appear and he would even venture a an answer somebody said white somebody said black somebody said blue the I've seen the marking scheme it says if you shine blue light onto a blue object it will appear blue that's supposed to be the easy question here's the harder one if I have a red object I now illuminate it with blue light what color will it appear a unanimous black fantastic I guess you understand that red should absorb a red object should absorb the blue light leaving nothing to reflect and that indeed is what the marking scheme will give you two points for but it's wrong on two counts oh if the thing was shiny it would reflect some of the light there you go and we call that specular reflectance so you could argue that I'm just seeing the light itself but you know the thing is reflecting that light so you could have a red object that kind of looks blueish because you're seeing the reflected light in it a mirror reflection what about this it could actually look great because we have a thing called fluorescence I'm sorry I'm gonna go back to this demo yet again because it's such a good demo this one can we get the lights down again this is the last time that we'll mess around with this I've got here agree a yellowy object it's a higher visibility jacket so I've got yellow light coming onto a yellow object and it looks yellow okay what if it was blue light falling onto a blue object a yellow object oh my goodness it still looks yellow in fact it doesn't really matter what put this under it will be a yellow of some type it really doesn't matter and down here we have high-energy blue light losing some of its energy in the chemicals inside here and reradiating lazier lower energy yellow part of the spectrum we call that fluorescence makes measuring an absolute pain because you put one type of light on and you get another type of light off which seems just stupid and not in the marking scheme what makes color well there's a great book called the physics and chemistry of color the 15 causes of coloreds by kurt Nasu pretty much each chapter deals with a different reason for creating color and as a physicist I was quite upset by this book because it all seemed to be chemistry a lot of it is to do with electrons and how they move how they absorb and how they radiate energy but there's that's only half the book there's the other half and there's some physics creeping into the later chapters let's look at some of it in practice in autumn leaves go through a series of incredible color changes all to do with the chemicals inside now normally you'll have chlorophyll in a leaf very complicated molecule that that wiggles some frequencies and absorbs energy it's using blue light violet light and red light collecting energy at these wavelengths but it's reflecting the green light not using the green light at all so we just see what the leaf doesn't need as the leaf dies we then see the carotenoids now chlorophyll is actually quite a weak molecule it'll fall apart quite readily so they have carotenoids in there too a mop-up free radicals that would otherwise destroy the chlorophyll so those chemicals are in there to keep the thing going they're also the reasons why carrots are carrot colored I guess there's a connection with the name and finally as the leaf dies and the chlorophyll stops working and a carotenoid stopped working we're left with an through silence and through silence a set of dyes that change their color with pH from violet to pink and it's no coincidence that the names violet and pink are color names but also flower names so how do we see color I'm going to do a human experiment on you now can we dim the lights a little bit James can you see the dude are in the middle of the elephant yeah stare at the center of that don't look left right up or down just keep staring at there okay and in a moment I'm going to remove the image blink a couple you're gonna blink a couple of times and then you're going to make an odd noise are you ready blink a couple of times I can hear the are noise that's good that's perfectly normal different parts of your retina are getting bored at seeing those colors and so they switched off they literally became blind to those colors and so when you see nothing you see the opposite of that color again I've got a different image stare again at the center keep staring don't look left right or Hort down we don't want that image to be blurred I want to to burn this image into your retina in a non blurred fashion I guess you recognize the shape don't look at the shape just stare at the thing and I guess you can see where this is going okay blink a couple of times it's wherever you look it's on the ceiling it's on your neighbor's face it should go after a day or two can you tell me what what the blue turned into yep can you tell me what the yellow turned into blue yeah you tell me what the black turned into yeah every color has its opponent color an opposite color the evil other half and designers know about this color there they go you can't get too you can't get any different between this color and its complementary its opposite color the scientist Maxwell use these afterimage colors to analyze how people saw colors and people with different vision defects color blindness see different after images and from that you can work out how we see color I did this talk in Belgium I felt I should use the Belgium flag it didn't work so well for me the audience were all going are but I just I just didn't it took a long time for me to see the colors and that got me thinking it's not just what you're seeing is what you think you're seeing it reinforces what you see and a little bit later after after doing that talk I was walking down the road and I saw someone with a leather jacket on with the Union flag on the back it was late ish at night and from a distance with the ill illumination I wasn't quite sure what I was saying but could pretty much see the color but when I got really really close I found out that there were no colors on there at all it was just in black and white and I was imagining those colors you can see things that are not there to remember that first slide we're working slowly towards that color is quite complicated if we look at the response of the human eye there it is and it looks like someone's thrown a psychedelic bedsheet over some furniture when they have there are some details missing let's remove that and look at the details we have three types of sensor in the eye responsive to different parts of the spectrum by different amounts there's one line there that shows that something is responsive more to the blue end of the spectrum than the others there's another one that's more to the red end than the others and there's something more responsive to the middle part of the spectrum and these three different signals from this mosaic of sensors in the retina gives us the sensations of color color is a three-dimensional perception and this is a machine that you can use to investigate how we perceive color this was built by my boss Julie in the 1980s and there she is doing some experiments she was repeating experiment that were done in the 1920s by wdw as he was affectionately known and John killed and this is what they did to see light to see color you need light in the first place and that accomodation level will affect how you perceive light so we set a sort of general field at daylight viewing conditions we took a spectrum now we grabbed just part of that spectrum in this experiment we're taking the cyan part of the spectrum and then we asked observers to tell us how much of three primary colors selected from different parts of the spectrum are needed to match that color can you can you imagine in your mind how much you'd need of blue green and an orange red to make cyan let's let's have a go at doing that if we had equal amounts of the first two primaries are no red we'd actually get a very poor match so observers in the 1920s and then the 80s were allowed to twiddle the knobs and adjust it to create a decent match for them and it's not doesn't always work first time but eventually you'll get a reasonable match now can I tell you that the actual settings on those dials will vary by person by illumination level by eye but you can end up getting an appreciation of how individuals see or sense color and in the 1920s those data were combined to create what we now call the standard observer that's the data all pooled together and we created a color chart based on the human perception of color there it is and in 1931 an International Committee agreed that this is the chart that we will use from now on and today that human model based on those seventeen observers is embedded in cameras and projectors and television and iPads and any color rendering system even traffic lights oh I did say that color is a three-dimensional thing but this is a two-dimensional plot we're losing something aren't we there aren't three numbers being represented here what we are not representing is the lightness of the thing so this is it's sort of a slice through a piece of psychedelic ham with lightness coming out of the paper you can imagine the third number telling you something about how bright it is anyway traffic lights there's a British standard about what the color of traffic lights should be there they are the Reds should be in that little red area down the bottom the greens there and the Amber's there why because we don't want people to confuse the different colors now you can actually plot color blindness color confusion lines for people with color defective vision on this chart let's pop them up there right now there they are and can you see how everything along that each one of those black lines that would be confused as being a similar hue for someone suffering from due to nokia a type of colorblindness can you see that there's no line connecting the red with the green can you see that so that's great those people suffering from that kind of colored vision won't confuse the two types of traffic light but there are three types of colorblindness because there are three types of sensor in the eye and there are three things to go wrong and those are the other chance as well how do we measure color well we could take a photograph of the spectrum here we have some white light coming in and shining onto one of these gratings like I just showed you here and you can take a photograph of what comes off and you can compare the photograph with a spectrum with a photograph of a standard spectrum of a standard lamp that you know how much light is coming out at each wavelength and from that you can infer exactly how much light is present and then pop it through a model of the human eye and then you can plot it on that color chart alternatively you could do the cheap and trick a easier version which is just to measure the amount of green light the amount a red light the amount of blue light the same way that the eye does get out three numbers and again that would give you a reasonable approximation this is the way the cameras work but the color that you can see and so we can use that to measure creme brulee or whatever you like but perceived color is oh so much more than just amounts of bread green and blue this is the second part of my talk and this is why I basically stopped measuring color because it just becomes so complicated it moves from physics to biology and psychology and so let's do it another experiment on you by the power of PowerPoint I'm going to rotate a piece of white so that it obscures these little dots one at a time it's going to do that okay can you see so let's do it and while it's doing that I'd like you to stare at the cross in the center and something weird will happen keep staring keep staring it'll vary the experience with different people will be different but what a lot of people see is that initially you see a kind of green thing going around and then maybe you only see the green thing going around - weird I'm not gonna explain that one it's too complicated I'm gonna tell you where all these changes occur from though which so we have a mosaic of red green and blue sensitive cells at the back of the retina and exactly how they're distributed and the proportions of them will vary and that's controlled by your genetics someone took a photograph of the different types of cell in the back of the eye and colored them in with a felt-tip pen that they really aren't this color this is just to show you the different responsibilities of different parts of the eye different parts of the eye and different people's eyes and so this really doesn't mean that different people can sense color differently I built a machine to measure how shiny cats are because someone wanted me to and when I did that I realized that we don't see light linearly different equal steps of increase in light don't appear to us as equal increases in light as perceived we have a very nonlinear response to light so there are got the whitest thing that the projection system will display and the darkest thing that it will display labeled 100 and 0 where do you think 50 will be let's put it in does that appear to be hot exactly halfway between the other two no again we could do experiment so you can start messing around with it but we have a nonlinear response to light because we have to accommodate such an enormous range of light levels age there's my dad and my daughter many years between now my dad's eyes have been radiated by ultraviolet radiation for a very very long time so they've yellowed with age which basically means he's not getting so much blue light going to the retina so for him if I divide the picture on the left hand side he'll see everything literally through sepia tones whereas what my daughter will say thick blue things a lot brighter light level we have another type of sensor in the eye responsive to low light levels which actually gets more excited by the blue end of the spectrum so if you go into the the blue Bellwood later dusk or dawn there's a reason to do it you'll see that bluebells look suddenly a lot lot brighter and yet they look less colorful so the way we see color varies with light level a lot if you ever done decorating did the color come out the way you wanted it to it's two things going on at least one is that as the light comes in it gets mixed around and so it will affect the other colors in the room the other thing is that when you're looking at a color swatch there's you're seeing an area that's different to the whole room being painted that that color the majority of the color sensors in the eye are located in the center two degree field of vision that's a thumb at arm's length that's where you're most sensitive to color beyond that there's not so much sensitivity so depending on the area or the thing you're looking at you will literally respond to the little color very differently so I had a I did a kind of mind experiment what do we actually see what's actually going into the brain well the first thing is that the eyes are pinhole cameras other things upside down secondly there every mark in 10 degree slots there's a blind spot I've marked that and black in this image where most of the light well it's not being received at all because you just don't see it that the cabling carrying the signals to the brain obscure it and then also can you see there's a little white disc somewhere in that picture that's that 2 degree field of view that's that's the bit that there's most color sensitive so really really really that's what's going into the brain what an amazing fix-up job that brain is doing for us it remembers what's going on for fun we are used to exploit this chromatic adaptation property we at MPL for our work we had a big machine for measuring light bulbs big sphere that you could walk inside and for fun we used to put a green light bulb in there people would go in there be amazed at the fact there were no ceiling floor or wall and the acoustics were great and while they were in there enjoying it or their their eyes were getting bored with the color green so when I came out everything looked pink color constancy is an odd one when you look at a scene you refer everything to the whitest thing in the image so here I guess it's that some of the lettering and things look greener or yellow or blue err than that and the eye is very good at adjusting so again the banana in this scene looks the yellowish thing and the orange looks orange but in an absolute sense in this image the grape and the orange are actually identical in color and the two color blocks down the bottom are actually identical in color so the eye does some critic odd things or the brain does some pretty odd things and in fact we have things called special color cues so bananas always look yellow independent of what color they actually are some great titles of scientific papers if it's a banana it must be yellow the role of memory colors in color constancy nearly finished I've got a few more little curious demos to show you this is such a good one it got nicked I had to take a photograph of it there's two bits of pink look the same to you they kind of do kind of don't the mall you stare at it maybe they do which is interesting it's all dynamic it varies with time because of what their next two colors affect each other and in fact those Pink's are identical but and if you stare at it long enough they appear to be the same and then if you look away and look back again that the image comes back the effect comes back and it varies with field size so if I do that they really do look very different indeed now don't they artists know about this they exploit it so van Gogh used to have green next to orange because they made the orange look more orange in the green look more orient more green and yellow stars and blue skies so the yellow looked yellow and the blue look bluer this is a really weird one the area's labels a and B do they look the same no they are if I put that there I've got a block there and the area labeled a right next to it looks the same B that's right next and the thing is a uniform in itself so it really are identical in color would you ever would you its craziest no three things going on here B is in shadow things in shadow looked lighter than they really are we open up the likeness of things in shadows B is surrounded by dark things so it looked lighter than it really is and the third one this is the weirdest this is the one I worry about most you recognize the pattern you know that B is light and a is dark and what you know is actually influencing what you're seeing is that mind control is that scary if you wanted to get machines to see color and light the way that we do you'd have to program all of this into them you have to measure the contributions and the effects you'd have to get the machines to recognize the patterns and measure things that are around the things that you're looking at this is my last slide it's quite complicated slide I'm gonna move that there can you see the green next to the green is green but if I shrink it down we have it does it still look green it depends where you're sitting depends how old you are depends on those depends on oh so many things all right this one it is the same green the only thing I'm doing is changing its size but does it still look green does it look the same green so depending on where you're sitting I mean that might look a bit like that that might look a bit like that again we can tweak this for personal taste it gets worse look at the eyes do the eyes at the same if I do this oh maybe I said those two the same um maybe the eye is a little bit darker but it's does it look cyan --is-- you or does it look gray you can see where this is going those two eyes are identical so we see these identical this is where I began this is all going to end thank you you

Channel: The Royal Institution

Views: 91,369

Rating: 4.9224019 out of 5

Keywords: Ri, Royal Institution, colour, physics, chemistry, optics, refraction

Id: af78RPi6ayE

Length: 36min 39sec (2199 seconds)

Published: Wed Jul 04 2018