 Okay, so we'll talk about the first steps in seeing, and we'll talk a little bit about cortex today. As I said, most of what we'll be talking about in this class, although the theme is the Neobiology of Consciousness, most of what we'll be talking about will be the Neobiology of Conscious Vision and Unconscious Vision. For a number of reasons. A, we have highly visual creatures. If you get a cold, you know, you don't smell, it doesn't, I mean, somewhat annoying, but it doesn't really interfere too much with your daily life. Well, let's say if you go mountaineering and you become snow-blind for two days, you're devastated, right? What you can do is severely curtailed. So we are, and you know, between 30 and 40% of our brains, 30, 35% of our brains are given over to vision and the analysis of visual images, the visual information coming in through the eyes. It's a significant fraction of our brains dedicated to vision. And there are many visual illusions and you can manipulate these visual images very easily and we know a lot about vision in animals. And we just know so much less about some of the other modalities and some of them are much more difficult to manipulate, have a slower time scale, etc. So that's why we're just opportune and focusing on vision. And one of the basic lessons of today's lecture is that you don't see with your eyes. What I mean to say with that is that although you need your eyes to do normal vision, not for every vision, right? I can close my eyes, I can still do imagery and I can still dream, all of which without my retina, you know, but for normal vision I need my eyes. But my eyes is not where visual consciousness occurs. The activities, the state of the various neurons in the retina do not reflect accurately, do not reflect the way I see the world, the way I subject to see the world. So although it's terrible in partners, we'll see. It's the first steps. Perception happens in higher stages, which is what I express by saying you don't see with your eyes. Okay, so this is the stuff which you see. So this is a transmission microscope, an EM picture of a retina. It's a retina of a closely related species, rabbits. I mean, closely related as biology goes. It's not much different in a human retina. It's going to be a little bit thicker than a human retina. It's probably like 150 micrometers, something. Think of it like a credit card. It's a little bit thinner than a credit card. And it's on the order, I don't know, of something like this, five square, five to eight square centimeters. If you take the entire, bless you, the entire retina. Now the curious feature of this is, one very curious feature, is that the light comes from below. So what you have in the retina. Okay, these are the so-called cones and rods. These are the photoreceptors where you have a transaction of the photons, the light signal, the electromagnetic signal. It's converted at this stage into electrical signal. So this is where physics meets biology, as it were. Now this is at the back of your retina. What you have here, where we can quickly move forward, this is a typical vertebrate eye. This is the lens. So the image passes through here, then, well, the way it's formed here and form a sharp image, usually sharp, unless you wear contact for glasses, at the back of the eye here in the retina. So this is the retina we're talking about. Which is shin feet of neurons. Highly, highly beautiful piece of neural engineering. And the picture we showed before showed a zoom-in of that. And firstly, the image, the photons have to go through all of this stuff before they strike the photoreceptors at the back of the retina. So the photoreceptors at the back here. It's really backward. So the light has to transverse all these different neural stages, which for the most part are translucent to these wavelengths, until it gets absorbed. The odd photon becomes absorbed here in the photoreceptor, gives rise to an electrochemical signal. So very complicated set of biochemical reaction that ultimately leads to a change in memory and potential. A lot of known is about it. This continues to be an area of research. But at the gross level it's reasonably well understood how light is transduced into electrical signals. And then it has to percolate through all these different neuronal cell types until it ends here at the ganglion cells. These are ganglion cells. So the ganglion cells provide, they have axons. So these are spiking neurons. They generate these action potentials I was mentioning. And they generate, they have an axon. And a million of these guys bundle up together. You can see a little bit of it and leave the retina. So the output, so what's nice about the retina, it's a fairly simple, it's a laminar structure that is both horizontal as well as vertical connectivity. And the input is well defined. It's light of a certain, I mean very limited spectral bandwidth. And the output is also very well defined as action potential at these wires here that collectively are called the optic nerve. You have between one and 1.2 million of these axons that make up the optic nerve. And if any of you have ever dissected an animal in a prep course, you can easily see this little stalk that comes out of the eye. That stalk is, of course, that includes the myelin sheet that insulates and some other fiber that insulates the entire assemblage of wires. But essentially you have 1.2 million wires coming out of each eye. Now you have all these different neurons that are called amachrine cells and bipolar cells and horizontal cells and they're very different cell types. But we're not going to go into that. What is important is, okay, this shows several very important facts. So first of all, you probably all know there are two broad classes of photoreceptors. There are two broad classes of photoreceptors. One are called the cones, photoreceptor and the other one are called the rod photoreceptors. And they're related and they use similar biophysical mechanism to signal, but they have different sensitivity. The photoisomeric substance, the substance that allows you to convert the incoming photon into ultimately electrical signal has a slightly different spectral sensitivity. So you have these very small photoreceptors called rods. You have roughly 100 million of them, between 95 and 105, 110 million. So the predominant cell type in your retina are these rods. And they have this distribution. This is at the center. At the center of your vision, the phobia, the point of sharpness seeing, the one you usually use if you try to really look at something. There are almost no rods there. And then it peaks at, it has two peaks, a bimodal distribution to the left and to the right of the phobia. Here's another odd feature of the eye. I'll talk about it in a second called the blind spot. There's a location in the eye at which you have no photoreceptors whatsoever. There's literally a hole there. The hole is necessary because that's where the optic nerve has to leave the eye. For that location, you have a blind spot, you have no photoreceptors. For that location, in principle, there should be a hole. There ought to be a hole. And you can see if I have two eyes, the input from the other eye can compensate that. But it's a little bit strange to explain why if there is a hole, why don't I see the hole? If I take a CCD camera and remove some of the pixels, even if you remove a single pixel, it's incredibly annoying and you'll send it back, you want to get refund because the single pixel is always black, always white. So why don't we have that? Well, it turns out there's a very clever mechanism that the brain uses to compensate for that. And so it's interesting historically, the blind spot wasn't discovered until sometimes in the 16th century in France, I believe, which is a quite remarkable fact. Come to think of it, it's present in our eyes and we'll talk about it, you can detect it, but it really wasn't known, at least historically, it wasn't known until a few hundred years ago. So then you have the second photoreceptor type, called the cones. They are between roughly 5 million, so 100 million rods, 5 million cones, this is in each eye. And they peak at the center, the point of sharpness seeing the fovea, they peak and then they rapidly die out, such that at roughly 10 degrees, 15 degrees eccentricity, you just have very few of those. And collectively, I'll talk about de-mediate color vision. There are actually three types, three subtypes that should be called short wavelength sensitive photoreceptor, medium wavelength sensitive photoreceptor, and long wavelength sensitive receptor, or SL and M type, but everybody calls them sort of as an abused elongage, calls them red, green, and blue photoreceptors. But I'll tell you what, it's actually not the right way of talking about it, since one of these photoreceptors isn't really selective to blue, it's selective to the entire spectrum. So you have this, this is very different from a CCD or a CMOS camera where you have a homogeneous coverage, like if you ever look at, if you manufacture them or you look at a CCD or CMOS camera, you'll see constant hexagonal grid everywhere. Here, although you also have hexagonal spacing, certainly in the fovea, you have this very irregular pattern, and this is a unique characteristic of retinas. It doesn't have to be like that, but all animals have some sort of specialization that you have this very high density that rapidly falls off. The reason for that is believed to be that you only have, at the point of highest seeing, takes a lot of special resources, you need a very extreme high density of neurons there, and that for various reasons, that was too expensive to maintain everywhere. So instead, it apparently it was cheaper from an evolutionary point of view to have this uneven retina where you sample a lot of information at one point, namely your point of sharper seeing, and then you have the complicated structure that enables you to move your eye rapidly around. That seemed to be... You could do it like a CCD camera, that's in principle, you could have a retina that has constant density everywhere and it doesn't move. But for whatever reason, animals didn't choose to pursue that. Now, not everybody has this fovea. Some animals, some grazers have what's called... They have an elongated fovea, a visual streak, they have a visual streak, so their fovea is in point line, it's elongated. It makes sense if you're a grazer, like a cow or something, you really want to be... You have to worry about a gazelle, you have to worry about predators, so you really want to have sharpest seeing all of here. Some animals, I think some birds have two foveas. So this tends to be something that all primates have, all the cats and other animals have, mice, I think, also have. Okay, so you have this, the point of sharpest seeing at the fovea. So the fovea, just to give you an odd of magnitude, so this is like a degree or a degree and a half. If you roughly compute this distance, this distance, take the tangents and radians, you see, at least I did it for my thumb, it's like a degree and a half. So your point of sharpest seeing usually is like a degree and a half or even less a degree. That's really the point of sharpest seeing. And that's where you have this concentration. So within a few thumbs outside of this, it falls off at least the densities reduced by at least a factor of 10. Let me come to that. To first order, yes. Now, you'll see actually a real retina looks much more messy, but there's no large-scale asymmetry. There are very fine-scale asymmetries between up and lower visual field, but to a first extent it's radial symmetric. So this is really just a point of very, really only the central vision. That's why you constantly need to move your eyes. Certainly if you want to read, it's very difficult reading things outside your... in your phobia. And this mediates color. So a number of things. So as I said already, this doesn't occur with our perception experience. We have no hole in our perception, although we have one in the retina. This also says this distribution tells us that the cone photoreceptors, which mediate color, sort of pita out very rapidly, but it's not that outside here there's no color. It's not that I only see here color and everything else here is grayscale. The world doesn't look like that. I can look at this and clearly that light source there sort of is bluish. So again, it seems to be... So you can argue that if you look at all these mechanisms that perception is this wonderful cone job, it suggests that everything is colorful, there's high detail and fidelity information everywhere, and why in fact this high fidelity information is only present at certain locations, particularly at the phobia. Now, these photoreceptors, these are the cones. They work relatively rapid. They have a time constant of, you know, tens to 100 milliseconds while the rods, these ones, work much slower. Now rods have... there's a trade-off there. The cones work faster. There are three subtypes, so you can get differential spectral sensitivity, but they also take higher... they don't respond as well to individual photon. Their threshold for responding is considerable higher in terms of, you know, if you just vary the light intensity and you measure when you get a significant electrical signal, their threshold is considerable higher than the threshold of rods. It's been shown already 50 years ago that if you put a person in dark and adapt them for 20 minutes and they're total dark, and then you release photons from a very weak light source where you can actually recognize that their individual photons that are being released, the dark adapted retina can pick that up with more than chance probability. In other words, your system is so sensitive that it's ultimately limited by physics. The entire system, this entire retina can pick up statistically individual photons. That really means you're limited ultimately by physics. And that seems to be a journalism in the nervous system that all these sensors are really by the physics of what is possible. Of course, you could lower, for example, what you could do, you could lower your temperature, go to liquid nitrogen, but that's clearly not a solution that's sort of easy to adopt for biological organism. So the argument is, given the existing neural hardware, the neural hardware does as well as it possible can, given the physics. Now, I mean, when would you like, based on what I said just right now, to see the rod receptors? When do you think they're needed in normal life? Yeah. So, I mean, not total darkness, but, you know, if you turn off the light here and we only have these faint room lights, or maybe one or two of those room lights, then you would use these rods. Certainly, when it's a starlight. Moonlight, probably, it's already pretty bright, but certainly starlight, you know, that's when you would use these receptors. It's a little bit... I've never seen a clean explanation. I mean, you have 100 million of them, but you only use them, you know, for a limited amount of time. Of course, you could argue it's critical at that point that you have this low-light vision system. So think of it like a low-light vision system. Certainly, under this light, the rods are totally saturated and you couldn't use that. We know that also from... there are people who are called achromats, chroma, color A, absence, achromats. So these are unfortunate souls who are born without any cone receptors whatsoever. Their eyes has no cone receptors whatsoever. And Oliver Sacks, the neurologist, he wrote a book about him, The Island of the Colorblind, that's in the Marshall Islands in the South Pacific, and there was a king who had a harem and way back in the 17th century, most of the island got wiped out by a tafoon except the king in his harem and the king turned out to be an achromat, so he passed on through his harem. He passed on his genes for achromating. Today, like 25% of the population, at least one Oliver Sacks, is that 25% of the population descended from the king and has achromats. So they have to wear either two sunglasses or very, very, very heavy sunglasses and they're only comfortable really at night because otherwise they just see... I mean, here they wouldn't see anything else. It would just be all bright and bright light because at these intensities, you can't see anything saturated and you cannot derive any signal from them. Also, they don't see any color. We'll come back to that in a second. So you might have noticed if it's dark that the way you really... if it's really dark and you don't use cones, the point of sharp sensitivity sailors will tell you this, or climates when you climb at night or when you sail at night that you really have to look out of the corner of your eyes. Why? Well, because at the fovea itself there are no rods, there are only cones. You don't have conditions, you don't have any signal. So you really want to make use of the... at the... at the temple side or at the nasal side. That's where your peak photoreceptor distribution is. That's why you want to look at dark a little bit out of the corner of your eye. Okay. Yeah, so these are the long, medium and sharp wavelength sensitive photoreceptor. Here what people did, Williams did in Rochester, used special camera, funnest camera to pier into the optics of two living retinas of two living, human, I think, male observers. And so these are... and then he sort of... he added synthetic coloring. So these are... you can distinguish by the absorption, you can distinguish the three different cone receptor types. And so you can see this is... I don't know, this is five arc minutes. So they're 60 arc minutes in a degree and a degree is something like this. So these are fairly small patches. And what's remarkable, these are from two different volunteers. A, the totally different one to the other. And B, even within one, it's highly... it seems to be a highly random distribution. It's not at all like... again, if you see this in a silicon retina or a CMOS, you would have very regular... you know, you would have the three types of filtering on a very regular grid. It's nothing at all like that. It seems to be statistically... so here three quarters of them, one are the long wavelengths, sort of conventionally called the red. Well, here only half. 20% medium. Here 44% medium. In both cases, you have very little short wavelength sensitive. That's to the intensive... to the high end, the intensive, remember E equals h nu. So these are the more energetic. This is towards the blue. You have very little blue. This was already known before. But what's remarkable about this... and again it doesn't accord with our... with the way we perceive the world that things seem to be totally inhomogeneous. See, even a little patch, it's not that you have a regular red, green and blue, but you have this very uneven thing. This very uneven distribution. And somehow your brain has to compensate for that. Because if you look at things something like this here, my nice colorful red vest, you know, if you look at just the red here, it seems to be relatively homogeneous. Or if you look at a red painting of someone, you know, if you go to... if you look at a cubism monochromatic painting, they seem to look relatively homogeneous. You don't have the perception that you have this highly inhomogeneous sampling of the world. Yet that's what's present in our retina. So again, the brain has to... the brain, after the red... not the retina, because this is what the retina sees, the brain has to compensate all of that. Now, the difference between the two is just an expression of the fact, and I think we'll see that more over the next 10 or 20 years with the completion and the ever-deepening of knowledge coming from the human genome project, that each one of us, I mean, we are scientists, particularly if we study animals, we always stress the commonalities with animals, and with respect to commonalities of animals, the variability among humans is tiny, compared, let's say, to shrimp that has 12-cone types, or compared, let's say, to male-new-old monkeys that have two-cone receptor types. You know, I mean, in fact, most mammals only have two-cone receptor types. But then what you see that if you look in detail, I mean, that's true, but in detail, there's a great deal of variability among us, and that variability will also be expressed in the future, and we'll be studied at some point. So here, clearly, I mean, to me it's reasonable clear that those individuals will not see the same the world in exactly the same way, because, you know, this guy has much more red. You know, if you look at, yeah, this is long, medium, so this is green versus red, you know, the ratio here is three times a third of the ratio here. You know, so one with a supposition would be that clearly this guy is going to see colors slightly different here, not totally different, but, you know, this guy will probably see red very different or somewhat different than this guy will see red. And we can begin to connect that now to the genome, right? We can begin, and people do that now, that they relate a specific aspect of perception, for example, red perception. We know that in man, there are actually two types, two different alleles here for the photopigment in those long wavelength cone photoreceptors. They're two slightly different alleles. There's a difference in a single amino acid, and it gives rise to shift in the peak of the spectral sensitivity of three nanometers. So it's roughly half-halves, like 55% of man versus 45% of man. In other words, roughly half of us guys here see red slightly, you know, have a slightly different photoreceptor, just have a slightly different spectral sensitivity than the other half of the guys. Now, and that's bound to, and in fact if you do very careful tests, that does show up. And I think there's going to be more and more of that that we realize that we each born with statistically the same sensory apparatus, but in each one it really depends on genes and on our environment. Okay, so not, as far as I know, not this one, although I think that's what... I think that's what they want to do. Well, so as I just mentioned, they have done it with respect to these two red alleles in man, so you know, you can directly test it, if you want a little bit of blood or spit or something, and then you can see which allele do I have and which one does the next man have, and then people have tested their slight differences in the red, as you would expect. So, yeah, so in general it's now a research. There's, for example, evidence now that women, that many women have seen figures up to 40%, many women have actually four different cone photopigments. See, how does it work? Okay, we all have, so you might know, let me see. How many people are colorblind here? Is there anybody who's colorblind? I mean, there must be, statistically. You know which one? Are you missing the red or the... Okay, so if we look at the photoreceptor distribution, let me see. Screens, center. So, okay, so this would be the long wavelengths, this is short, this would be sort of towards red, and this would be towards blue, and I think the rods are somewhere here. I don't, I think, okay, I didn't include a picture here, but I think in terms of sensitivity, something like this, the most sensitive, as I mentioned, by two other magnitude are the rods. And you have only a single rod type. So, clearly if you don't have cones or if they're inactive because they're not sensitive enough, then you can't perceive color because you only have one type of photoreceptor, so every signal, whether it's a lot, you know, whether it's a, you know, if you think of monochromatic signal, whether it has, you know, 400 nanometers or 800 nanometers, it's going to be absorbed to some extent, depending on this curve, and this photoreceptor can only signal, there's only one way it can signal, it can only output action potential, and this action potential you don't know did it come because there was a tiny bit of monochromatic light at this wavelength or because there was a lot of light at this wavelength. If you have a single photoreceptor type, you cannot tell. That's why achromats cannot tell color because they only have one photoreceptor. It's a very important point to remember. You only have one photoreceptor type, you cannot pick up anything differential. So this is an instance of population response we talked about last time. The only way you can pick up color sensitivity is by comparing, let's say, this output to that output. Anyone cone, anyone photoreceptor by itself and if you had, let's say, only this one, again, you couldn't see color, all you would see is intensity. So once again, it's very important to remember if you're thinking of the action potential as an electrical signal and you're looking at this neuron, you don't know did it just give me a signal because there was lots of this or lots of this or little of this. All you know is the neuron fires, so all you know is there's some intensity there. So color you only get by differential comparison. And in some men, let's see, either this one or this one is missing because they're on the sex gene, they're sex links, that's why you see it much more frequent in men than in women. This one seems to be present in almost all, this one is very rare not to have. It's totally separate coded on a different chromosome. So either some, roughly seven or 10% of men either miss this one or this one. Now there are two alleles of this, so actually there's this one and there's this one. And this differs by, I don't know, two to four or three nanometers, something like this. Now some women express all of these cones for the receptor pigments in different, well, at least they have the gene for it. So that's uncontroversial that some women, particular women who have colorblind sons, and I have to remember why that's so, but particular women who have colorblind son will have genes for all four pigment, photo pigments. If these get expressed by the retina, and if they actually functionally wide up to the brain, that's a big if, that's not known yet. So I know it's still controversial, people are doing research on it, it's still controversial, then it's possible, then these women could discriminate colors that to their other sisters, their normal sisters with trichromatic vision and all men would look identical. That's how you can test. How can you test that men are colorblind? Well, you show them two colors that to a normal person with trichromacy with three color vision would look different to a guy with only two cones that would look identical. So likewise these women, they could tell apart that things that look the same to me are three color cones that would look different to them. So, yeah, I have to think about that. I know that's a fact, I don't know why. I don't know. I mean, that's how they look for them. That's also how they look for them. They advertise in England, they advertise women who are colorblind, then they do these color tests. As far as I know, the latest paper, it's still controversial, so I think what's not controversial anymore that these women exist that have genes for all four, I think what as far as I know continues to be unclear whether they actually express it, whether the brain actually makes use of this additional information. So again, what I find fascinating is that you know that you can A, tells you a lot about specificity, that this tiny one base per difference in the sequence coding for this protein can show up in psychological measurements and also that humans don't really see the world exactly the same. They see it similar, but they don't see it exactly the same. Okay. So I think I covered most of this. So the important point is the third one, that any individual cone will have some sensitivity, right? As I showed you this curve, I'm not going to show it again. You can change monochromatic, you can have a laser monochromatic light source and you systematically change the wavelength of it and then you find at the same intensity there's some differential response. But that by itself cannot be used to compute color. The only way you can compute color if you have two or more photoreceptors with differential response and then you can compare this against this or you can do it with three or with four. And that's how color arises. It's another instance of a population coding that's very common and very prevalent in biology. Yeah, and the other point, there's nothing sacred, there's nothing sacred about three cone types. You can show computationally that given the natural distribution of light sources, which typically is the sun in the natural condition and given the reflectance function of objects in natural environment that you capture most of the variability of brightness and changes in wavelength with three cones. But it might well be that if you really care about one particular type of food because that's really what your species eats then you want to have additional color resolution there and then you have additional photoreceptors. And I said some shrimps that are very small retinas but they have much more color selectivity in certain frequency bands because obviously it must be important for their survival. So in most of us it's trichromacy but in some you only have two or in some cases you have sort of no cone receptors you only have rods and in some you sort of have some sort of supervision. So how do you test that? If you're looking of course at a computer monitor you're using your cone signal so the only way to test that is really go under very dim conditions I mean at night most things look gray if you really go out in the forest at night if you're hiking or climbing anything things look very gray. You can tell different shades of gray and when there's moonlight then you can sort of things acquire a little bit more color but if it's really pretty dark you don't really see a lot of color. So I think if you do see that still it's probably must be top-down effect that you know you might have some knowledge that I know my car is blue but it sure doesn't look blue and I mean it might look blueish and I suspect that would be a top-down effect it certainly doesn't look blue at night but I mean you have to be sure you are actually under hot conditions. Sometimes what I've noticed at dusk I can sometimes like if I look at a garden with flowers sometimes things look to me like I get additional color resolution and I think what might be happening is that at that cusp when there's still some light the cones but the light level is dropped sufficiently so now that I get a signal out of my rods that I then actually have four photoreceptor types that I can use and that gives me additional color resolution. I don't know whether anybody studied that it's just, at dusk sometimes things look very vibrant they seem to acquire this additional this additional color and it might be possible that that's involved there. Yeah so there's so color is you can argue that color is already very much a derived property that it's not an original property what I mean it's not primary property so you can say for some depth we all have depth perception I can estimate how long the distance to my hand I can estimate the distance to her etc and that's an absolute distance two meters or whatever it is or motion I can estimate so many degrees per second it's not moving with respect to me color there is no such thing really as color in the world what you have is continuous light source distribution you got something like the sun that radiates of course in a huge spectrum okay that's that reflection strike surfaces these surfaces will reflect absorbed light in different proportion as a function of frequency and then that gets reflected into my into my eyes and we as human we then map it onto these for most of us three dimensions somehow and then of course we have these names these color names that introduces additional level of processing but all of that is rather to certain extent arbitrary and as I said other creatures or even humans can have more or less color discrimination so I try to think that color in that way is different from a lot of other things that seem to have a more objective standing like distance and motion or surface orientation this is either zero or 90 degrees it has an absolute objective orientation but it doesn't really have any objective color although I would say this is black you know chalk on it okay so then there's this other thing in our retina I mentioned already the hole okay now you should just try it you probably haven't done this since grade school so just bear with me you should all do it so just try it again just to assure you so close one eye don't close both eyes you won't see a lot close one eye and fixate so it's on the temporal side on the temporal side so on the temporal side at the horizontal somewhere here 10, 12, 15 degrees my finger would totally disappear I mean the tip of the finger it's quite noticeable if you move it in you can see it and if you move it outwards you can see it 2, 3, 4 degrees would just disappear you should all just probably haven't done it in a while you too Patrick okay can you all see that I mean can you all not see that so I mean what I find mind-boggling and I'm not convinced this is true but the claim is that people didn't discover this till you know sometimes in the early Renaissance I find that difficult to believe I find difficult that the Greeks for some didn't know about this but apparently there's no historical record I mean maybe somebody did but there's no historical record even you know I was totally it didn't it doesn't talk about it so this is an indirect way of picking it up you notice it by the absence of something but there's not a hole there so what seems to be going on there's an active interpolation mechanism the brain actively interpolates actively does something to fill that in it does not just have an empty it's not like the back of my head where I just don't see anything now you can check there is an interpolation mechanism you can check you can take focus on something like a pencil you should try that so first of all you move this works particularly well with this orange head on it because I can move it such that I don't see the orange anymore there I don't see the orange anymore quite remarkable and now what I do but if I move a pencil across it the pencil seem to be continuous so if you move the pencil so let's say the blind spot is here if you move the pencil right over the blind spot it's continuous there doesn't seem to be a hole in the pencil okay so that's a bit funny we know you don't see anything yet the pencil doesn't seem to end there so what seems to be going on that the brain takes that the brain takes information in the neighborhood of that from the same retina and sometimes even from the opposite eye of course here we ruled it out by closing one eye so it also has to be from the neighborhood of that and then interpolates it fills in okay what you could also do in principle do and I think some people want to do that for cameras that have high high where you have low yield and high pixel counts one way is to get rid of that hole is to not to replace the silicon which might be very expensive but to actively fill in so you can say well if in the neighborhood of that pixel everything is red then I'm just going to assume that pixel is also red and I'm on the output of my camera I'm going to say red that's the brain does something like that it fills in based on property in the neighborhood so here what you have you have a nice vertical edge above the blind spot and a nice vertical edge below the blind spot so what the brain does it paints in it fills in this picture and says well probably what you have is a straight edge and that's what you tend to see okay but so again if you look at the output of the retina the output of the retina there's no information at that location this is something that the brain has to fill in what is this writing on the paper what is the brain going to like interpolate in the gap and turn those tears right a continuous what do you have different words in and on the spot you have like some other words so what is this brain going to interpret okay so people have done experiments not with words because I don't think you have resolution out there to read unless you make them very big but people have done different experiments this guy Ramachandra he's written an entire signing American article where they do where they sort of try to investigate how clever this interpolation mechanism so they've done experiments like that where they manipulate things separately in the horizontal and in the vertical direction those experiments have been done from psychological point of view and you know it's a limited it you know it's a limited mechanism it's not all that smart but it seems to work so well well that for hundreds of years people didn't realize that they had a hole there it's claim it's also there's a claim that this might also operate all the time for example there's this guy Kenneth Craig who in the thirties actually volunteers a psychologist that came to England he burned a hole in this retina by looking into the sun I mean he did this on purpose dedicated scientist don't do this please and the claim is that if the hole is very small after a while that the brain uses an active interpolation mechanism to compensate for that this might be a mechanism that's present all the time not just in the for the blind spot but is present all the time that if something gets damaged at least if it's a if it's a small hole that gets damaged the active mechanisms that try to compensate for that which would be of course a fairly clever thing to do for the brain okay so the retina I didn't mention this because we don't talk about we don't love it all but the property of the retina and their hundred and twenty one hundred thirty million neurons most of the neurons signal not with these action potential those pulses I mentioned last time but actually signal with continuous changes in membrane potential and it's only in the in this part of the the proximal retina that that neurons change from a continuous code where events electric where visual events are signaled by a steady D or hyper polarization by the time you get to the last new one here these retinal gangon cells you have the conventional action potential so this is an accessory called signals by the time you get to the action potentials by the time you get to retinal gangon cells the output action potential the reason is they have to travel a long distance they have to go out of the eye into the other structures of the brain the colliculose and cortex etc and that's a long distance you want to signal information very quickly and that's the way how the brain does it but within the retina when you have very often the brain these neurons use greater potentials but now what people have done you can record either by putting an electrode inside the eye or by recording from the optic nerve outside the eye this can also be done in humans this has to be done in some human volunteers incidentally who are blind in fact what they did they stimulated they put electrodes in here and they stimulated this yes I think the reason people say it's a developmental one it has to do with the way the brain is essentially developed from outside to inside so it's a developmental thing it's not I would say there's a main disadvantage because you surely must lose some optical quality because you have all that goo that stuff you have to use from and there's going to be some absorption so I would and I think the squid if I'm not mistaken the squid and octopus they have a direct imaging system so they have the retina the light strikes first the photoreceptors and I don't know whether people have ever compared sensitivity whether they are better per unit you know there's somehow higher sensitivity so no so it's purely it's a developmental one I don't think there's any advantage to it okay optic nerve yep I don't know maybe they cut off good question this in fact is even from an older paper they took it from older paper I don't know yeah you would think good point I don't know must have something to do with the way they filter their signal or they cut it off maybe it's a very high gain signal and they cut off most of it and this is excecella we call it I mean if you look at the ganglion cells I mean I've looked at the action potential they look principle node different than any other action potential so this is Schubel and Wiesel were one of the earliest explorer of the retina but particularly the cortex so this shows you this is based on an even earlier work by Stephen Cufflight at Harvard who sort of pioneered studying the retina in animals like the cat which is really very similar organized to to the human to the primate retina and what they discovered that most many neurons are most neurons this is a ganglion cell what is selective to to first approximation the radial symmetric spots of light so for example on the top is just if you just have black the neuron won't respond this is its background rate this is I think 800 millisecond the stimulus here so you know the neuron might discharge in the dark once a second on average something then you put a little spot you flash it on here it fires very briskly and then sort of it settles down to somewhat more the stimulus charge you have a big spot it fires less or then you put a spot but you have an annulus and then it's actively inhibited and then you have a different type of cell this is called an on-cell it's called an on-cell because if you put a little spot of light into parts of its receptive field so now again we come back to this notion that's so critical to the systems, neuroscience receptive field so here the modules were first defined in retina and they were defined in terms of space so you take an animal, it was first done in a horseshoe crab but the principle is the same, you take an animal you paralyze it or you get it to fixate and then you draw out where from which part of the visual field can you get can you excite the neurons so you have a cat staring at this black ball and then you notice every time you light flashes on here you can hear the audio monitor and if you do it over here the neuron doesn't care only if you do it in this region that's its receptive field and so here you have an on part of the receptive field when you flash a light in the center part of its receptive field it likes it okay when you make it at some point when you make it too big the response goes down again until if you make it really big it's the same as background okay and then if you you can actively inhibit it by flashing light only in its surround so if you flash light in its surround it's negative response you flash it in the center you get a positive response and now you can imagine if you do both then you get a sort of linear superposition right so then you get this which you can think of as that minus that which is roughly that and then you can ask to what extent are they linear etc but so here you find the classical receptive field structure early on in the visual system which is roughly circular I mean that neurons tend to be you know so you might have a structure like this and then something you know you know things are in perfect circles here but you have something okay that's not a very good rendition I was trying to get the fact that these are not supposed to be perfect circles but so you know you have a part of it so if you put lights here the neuron fires and then you have a zone around here that if you put things here light spots here the neuron will be suppressed and if you shine a light across everything here then you get some excitation some inhibition and now it depends exactly where the balance is so usually if you use large stimuli the neurons don't respond all that well if you light the entire visual field you might get a short quick burst response but then the neurons stop firing this is called an on neuron now you have the compliments called an off neuron where you have a central region where if you shine light it's inhibited and then there's a surround region where if you shine light it's excited so that's what you have here you shine a light in the center it gets inhibited and then once you remove it here you remove it the neuron fires and here the offset from here you shine the light on it's dark in the center the neuron fires remove the darkness the neuron stops firing now you can think of these as rectified you can think of this as a way for the brain to deal with negative numbers because how do you signal negative numbers if your code is pulses so if the key code to first approximation is pulses so neuron fires 10 pulses per second or 100 pulses per second how do you signal negative numbers one way to do it is you have two classes of neurons you have to rectify both are positive both signal with pulses one population signals the absolute of the negative part and one signals the absolute of the positive part and so you can think of these as signals that on and off that correspond to the half rectified positive and negative signals and that's one way you can get around signal negative numbers the other way to signal negative numbers it might be expensive you could of course have a neuron that fires all the time at 100 hertz all the time and then a negative number would be a reduction of that when you go from 100 to 10 hertz that's a negative number there of course that's expensive but the key is maintain discharge at 100 hertz some neurons do that but in the retina the way the retina seems to have chosen the many sensory system is to have this on and off strategy of having the positive signal carry here and the negative signal also carry by action potential in the off pathway and the point is these things are all elongated what these neurons like are elongated spots of light circles of light of various dimensions and one last feature before we leave the retina is that that if you go from the fovea there's something called eccentricity you can define a visual angle if you look at the retina you can see zero the point of origin where the fovea is and then you go further away from the fovea and you define in terms of visual angle and when you see what you can see what you can observe that if you look at the retina you find the same cell type close to the fovea in distance instead of degrees you have one cell type that systematically increases in size or you have another cell type that systematically increases in size so you have this nice geometric progression that in the fovea you have several cell types and they are very small they are very small than the retina this is where the synapses provide the input and all the electrical signals here are summed at the cell body you have the threshold decision whether to find action potential or not and that signals then output on that optic nerve on that axon that makes up the optic nerve that goes to the rest of the brain and here you have the same cell type in the distal periphery it's just larger and larger and larger it covers larger and larger parts of the visual field it makes sense of course this is also how you lose resolution because here in the fovea or close to the fovea you have one photoreceptor that maps into one ganglion cell and that only looks at a tiny tiny part of the visual field of information on the periphery you have maybe 50 photoreceptors that talk to this that feed all into this one neuron to this neuron obviously will have access to a much larger part of the visual field but its resolution of course will be consequently much less than the resolution of the neuron in the fovea this is from Dick Maslund at Harvard, MassGen in Boston and it's a very nice picture because now we're beginning to be at the stage where we're still very far away let's say from in Cortex on Thalamus where we can classify all the different types of neurons in the retina and I won't bore you with these different descriptions these are cones and these are horizontal cells bipolar cells, amachrine cells, ganglion cells so when I think ganglion cells it's not you know there are 1.2 million axons of ganglion cells these are not just one type there are many different types, they're on and off cells some cells as I showed you only carry the on signal there are two big additional channels of information these are called pathways or channels as I showed you before those neurons you have at least 70% of the neurons at any given eccentricity are relatively small and seem to be specialized in carrying information that's related to color to wavelength and related to high spatial high fidelity information that's 70% of the ganglion cell these are called pavocelar neurons after the Latin name for where they terminate the cells look pretty poorly developed they're called pavocelar neurons then 10% of the ganglion cells another Latin name, magnocelar they're big, well developed they're only 10% so there are many fewer and they don't carry wavelength information and they seem to specialize more in change information in temple information and this is a general lesson in biology so I'm showing you that you have different so these are all the output neurons only these guys project outside the retina their axons, you know, a million of them together make up the optic nerve but as I said, even at one eccentricity there's not just one cell type, they're different cell types and it's a general story which is then replicated at each point in the visual field that the brain they're all these highly specialists that, you know, for a job like color information or transmitting off information transmitting motion information there's a dedicated separate neuronal type and that does that so there are, I think, 12 different 12 different cell types now here with the 10 10 different ganglion cell types, there might be more subtypes that each specialize in carrying different information and the wonder of it all the miracle of it all how it all fits so well together I mean, I get up each morning and when I think about it when I'm not too groggy it never stops filling me with amazement because it's not like so this is a general lesson of the class what we'll discover there are all these specialized networks they're the special part of the brain some deals with motion some deals with color, some deals with hearing I mean, I told you there are networks that just do nothing but is somebody scared or is that person angry highly dedicated networks but it all fits together so effortlessly that I don't notice any of this I don't have 100 years of science to untangle all that secret it's all integrated in this wonderful smooth transparent interface talking about an interface this is really a beautiful interface so I don't know any of this all this complexity is totally hidden from me because you might imagine, well I might have to sort of scan systematically through these different channels now I'm looking at the motion now I'm looking at the color now I'm looking at the stereo information in a single interface and that's pretty nifty that's pretty nifty, no if you think about it and if you ever sit in a lab if you're interested in this seriously you should do you listen to neurons then you're even more mystified because those neurons they look so when you hear the cackle on the loudspeaker they sound so random to you I mean out of all of that excellent discharge firing has to rise all of this beautiful world ok so and the general moral is here that you have specialist neurons that seem to specialize for different modalities it's not absolute but they all have sort of one specializes more in color information the other more in motion information and all that information is sent out in parallel so you can think of a different from a ccd camera you don't just have one camera coming out of the system one pathway or channel the language people neuroscientists use but you have maybe 10 different channels coming out some carrying on, some carrying off information some carrying high fidelity information some carrying motion information and then there are whole bunch of subtypes about which we know much less that go down to special parts of the brain that are sort of mainly involved in boring house keep functions that are incredibly important for us but they're boring what I mean by that for example neurons that of course do nothing but control your pupil you know if you go now out in a bright sunlight you know your pupil takes some time to adapt they're all the special circuitry that does that's responsible for eye movement they're very very sophisticated, very fancy you know you're totally oblivious of it there's six different eye muscles all that has to be controlled well there's special neurons that do that the neurons that involve circadian rhythm the neurons in the retina gang cells it turns out that they are selective to the overall level they regulate that and influence our melatonin level etc these are all in addition specialist neurons that sort of that are involved in some odd little place in the brain that computational isn't very interesting but if you're impaired can cause you a great deal of grief so this shows here this is the two retina and so the predominant output in us and after the monkeys 90%, 9 out of 10 go to this intermediate relay station called the lateral geniculate nucleus it's part of the thalamus the thalamus is a big structure, well big it's maybe like a like a coil egg, something like this maybe an inch across, it sits pretty much in the middle of the brain, you have two of them left, right and these are pretty important if you don't have your thalami you're in big big trouble you're in serious trouble like you're in coma you don't see anything you might be alive but barely so and from there on so thalamus almost every modality bar 1, namely olfaction sends its effron output first to a relay station in part of the thalamus and the visual part of the thalamus the one that receives the output from the retina is called the LGN we'll talk about that, I mean we'll mention it a lot LGN and this then in turn projects to the back of the brain probably to the cortex so there's no direct connection from the eye to the cortex but you go through this one intermediate relay although it probably does much more than a relay but you go from the retina to the LGN and then from cortex which is not shown here LGN there are a number of things well we'll talk about it later and this is this pathway from the retina to the LGN to visual cortex that's the one that subserves conscious vision so we believe, I mean so people in general believe because 100% of the fibres also go elsewhere you don't really have branching it doesn't seem to be very common I mean again you could have branching you could have one neuron that sends its output both here and to all of these other brain structures in general in the primate that does not seem to be the case also there's no information that comes back into the eye that's different so in some animals like a fish for example for some weird reason I have no idea why there's an input from the olfactory bulb back into the retina okay so the fish sees with its olfactory brain no idea why but then fish are odd so in us there's nothing that goes back from the brain there are no neurons that go back from the brain back into the eye now 10% of the neurons go elsewhere so a big part of elsewhere is a structure at the top of your midbrain called the superior colliculus you have two of them you have everything you have two of almost everything it's called superior colliculi both of them and these are structures that in many animals like birds and reptiles and amphibians tend to do the predominant amount of heavy lifting in vision so before you really have cortex like neocortex is really a feature of I mean big neocortex is really a feature of mammals before you had that you always have the superior colliculus or the names these are sometimes called the tectum tectum for a roof because I guess it was sort of on top of the brain the roof tectum it's another name in primates it's called colliculus in fear colliculus that is responsible for audition superior colliculus is responsible for vision and you have two of them one on the left and one on the right and for us they seem to be mainly sort of you can think of them as evolutionary older part of the brain it's been there for a long time and it's responsible for things like controlling eye movement seems to be terrible important so people have lesions there typically they have various disturbances of their eye movements and then you have all these other for example you have the super charismatic nucleus super charismatic because this is a kaizen this is where the brain crosses I'll talk about that in a second and above that you can see very nice instructional imaging you have the SC and that's responsible for circadian rhythm so if you have disturbances of sleep wake cycle they tend to involve here that's also where the clock the circadian clock sits and like this guy gets input from neurons here in the retongue ganglion cell that seem to signal something about the background level of intensity and that's important for circadian rhythm and when things go wrong like in seasonal affective disorder and then you have all these other minor nucleus I just give you their abbreviations like that they are important you don't know about them and they do all these things like moving your eyes controlling your pupil diameter and other stuff so I mentioned already the two dominant passes are not only one so 70% of neurons are of this type 10% or 20% of this type and then they're 10% that also go on to the LGN and to the cortex proper that do other things that are less well understood but the two big population are called the power cell and the magma cell and as I mentioned one seems to be very sensitive to color they have small receptor fields small cell bodies and they seem to signal they seem to signal information about small spatial details probably they can do that because they have small receptor field but these guys don't seem to care about wavelength they have large receptor field and they really respond and they care about motion information the number of illusions that people have developed that we'll mention later on okay eye movements we want to talk about eye movements how many eye movements do you think you make a day for a year how many move your eyes how many move your eyes one time a second, ten times a second, one time an hour come on you mean out of magnitude trillions, what's that that's trillions, that's ten to the twelve that's a little bit high okay so you move your eyes roughly as often as your heart beats typically you move your eyes two to three times a second now that doesn't quite a call with our with our feeling, our perception I mean even if you draw my attention to it okay I move my eyes now but usually I don't think I move my eyes but if you actually look what people do this is what we did here, this is done in the lab this is a fractal landscape and this is an overhead imagery of I don't know Washington DC and we have people just look at it and this is the size of the box is how long they stayed at that location so we have these eye trackers and you can observe people now with high speed video cameras they constantly move their eyes all the time and decoding that tells you a lot about people what they're looking at, what they're thinking of in fact it's used, it can be used as a form of I mean we all use it implicitly to judge people we say well that's a shifty character because he never looks me straight, no we do that and I think it bears some truth like many of these things he never looks me straight in the eye you just don't trust a person like that and of course in some cultures you don't look people directly in the eye because it's a threatening gesture eye movements are terrible important in particular eye movements with respect to people and you can certainly, I mean I try this now you can certainly observe whether people for example tend to move their eyes to upper left or upper right there's some influence of whether they're very emotional driven at the moment and I think they tend to move it more towards the right and if they're more cognitive driven there's all these sort of interesting facts that you can study people they're sort of significant asymmetries in the eye movements and you can tell something about people by the way they move their eyes of course if you're trained actor you can control all of that but the fact is, and they're different tops of eye movements so the eye movements I talk about are saccades it's a Greek term okay I should put that in saccades, S-A-C-C-A-D-E so these are rapid eye movements, they can be quite big you know they can cover sort of 30-40 degrees although if they're really large people will tend in addition to move their head not all animals move their eyes for example if you look at rats or rodents I mean rodents in general they tend to move their head much more frequently than they move their eyes dogs are terrible expressive you know three dogs at home they can really sit there and you can see just like us they sort of track you and every time you go to the kitchen and all of that they have these very expressive eye movements and of course we tend to associate eye movements with humans that's the reason why I think we've chosen general humans like dogs because they have these expressive eye movements then there are also other types of eye movements they're so-called smooth pursuit where you smoothly follow this thing, you don't do it with the saccades, you do it smoothly in fact, and this is a very different eye movement system we know quite a bit about the new basis of this with smooth pursuit versus saccades and the other minor forms of eye movements when you try to focus on when you move in depth you have to control both eyes simultaneously so there are a whole set of different eye movements most of them you have no idea you make them and of course if I point them out to you you can be conscious of some types of eye movements what's interesting about eye movements is what's interesting about eye movements this seems to be a wonderful case in the months of a very sophisticated system that you can show, rigorously has access to information that the conscious you doesn't have access so you can show that there's disassociation when you, if I ask you do you see something change or not, you see no, I didn't see a change yet your eye movements are because it effortlessly moved your eyes to one new location so you can have these nice disassociations so Francis Crick and I call these zombie systems it's one zombie system so there are sensitive motor agents that are highly specialized to subserving one particular type of behavior in this case they subserve smooth pursuit or they subserve saccadic eye movements and usually the system itself does not seem to generate or require consciousness to work it works quite well in the absence of which is not to say that you cannot be conscious of these things but usually you're conscious only after the fact with a couple of hundred millisecond delay so the thing to say is that stop all eye movements, image tends to fade you can try this yourself it's difficult but you can try it so today our magnet yesterday our magnet arrived magnetic scanner here and so within a month we are going to look for volunteers to get your brain scanned and then because making even small eye movements we're going to ask you to have a bite bar so you'll get a plastic something which we mold up to your teeth to prevent movement of your head but to further minimize movement because we really want to optimize the signal to noise we ask you to minimize eye movements and then when you really try to fix it something when you really try not to move very often you get this fade out everything turns gray and what you've got to do you have to do this in order to restore vision you can also do this artificial so this you can try yourself it takes practice but you can try it not to move your eyes at all and after a while and you can also do this using electronic feedback that I measure at high speed a couple of hundred hertz where your eye is I position a picture on your retina and now when you move I sense that and I need to shift the picture at high speed so the picture always stays at the same location and then also very quickly you get the image fades you get this fading fading is not only retinal phenomena it's an interesting question because it could be used as a strategy to study consciousness because you can ask where does the fading take place is it a retinal phenomena is it a cortical phenomenon it's probably both and you have access does the brain have access to information that you consciously don't see yet some other agent in your brain still makes use of that those are all interesting research questions let me show you a movie so one interesting thing about there are lots of fascinating information about eye movements one is that you cannot see your own eye movements and so it's another case of a dissociation between what you see and what's out there so when you're looking in me you can clearly see my eyes move so it's not that I my eyes move so fast that you can't see them so I might move my eyes at 3, 4, 5, 6 if I really try and I'm young and all that I can maybe move at 600 degrees per second but you can still see that you cannot see your own eyes move you should try that when you go back to your bathroom just do this look in the bathroom mirror and look from here to here and you'll not see your eyes move what you'll see is your eyes at the beginning of it you'll see eyes at the end but you will not see the transition so what happens is there's a process called saccadic suppression saccadic, I move from suppression and it consists of a very different component masking all of that I don't want to go into it but the fact is you don't see your eyes move during that time yet you don't have black it's not that you see something there's a black out and then you see something again but again it's another interpolation mechanism where the brain fills in so it's like a dynamic it's like a movie you're missing let's see 80 milliseconds because that's you move your eyes and so the brain cleverly interpolates it takes something from the beginning when you fix it it takes something from the end and it interpolates that and then sort of it splices that into the movie of your life you can see one eye rotate if I focus my eyes looking in the mirror as long as I pay attention to you do the left or the right eye doesn't matter which, I can watch other eyes all rotate away you heard of that that's interesting but it only works when you de-focus so I have one eye moved still and then I can watch the other eyes in a different way so you can control your eyes independently you can do that regularly I mean you can okay can we use you as subjects? huh okay most of us can't, I mean most of us our eyes are always yoked I mean I have no idea how I could move my eyes independently I'm not sure I want to anyhow so it's the saccali suppression is an interesting phenomenon because again it suggests there's this disconnect what's out there as I said before we all have at heart these 90s realists we think what we see is really that's how the world is but it turns out you know even just in the last two lectures I think I've given you a whole range of phenomena where what we see is not really what's out there so what we should see we should see a movie the A that's interrupted well if we move our eyes and we do continuously it's you know if you take a video camera you continuously do this now that's gonna you know your viewers won't stand for it they'll throw up right so the only way to buy a video camera is to move very slowly because people hate this yet our eyes do it all the time we do it all the time we move it you know three times a second so that's probably one reason why people think you have this saccali suppression is that it tends to when the image is blurred in any case so you would lose high fidelity high spatial frequency information due to the blur you shut down the mechanisms are being sort of heavily debated right now but it is a clear case when it's partly shut down yet it doesn't look that way subjectively so there has to be another mechanism that compensates for that so I'll show you a movie that Laurent Etier a CNS student used to be in the lab did to make this point this finishes his first lecture and then I'll just briefly well no we'll just finish I guess today the movie that makes a point that while the retina is important for seeing that of course visual information has to be has to be generated you know at the photoreceptor level you have to translate photopic I mean photons into electrical signal and that's all necessary for seeing the nature of the electrical activity or the neural activity in the retina does not correspond to visual perception so A there is a dramatic decrease in spatial acuity although the world sort of when I look at it it doesn't look blurred it looks sharp everywhere provable I know it's sharp here and it's less sharp out there but it doesn't look that way yeah I failed to mention so in general you have three so where you have cones you tend to have three cones not equally distributed the short wave cones the blue ones always tend to be in a minority you always have many more of the long and the medium range ones furthermore at the point of very sharp as seen the central half degree so that's maybe like this you only have long and medium wave length you don't have any short wavelength the argument is it's an illusionary one it's difficult to really prove rigorously that you don't have that because you want to minimize chromatic aberration you get chromatic aberration and of course the chromatic aberration is stronger the higher the frequency of the light and so with blue you would get the maximum chromatic aberration and so the brain wanted to avoid that this is how the story goes the brain wanted to avoid that therefore at the central part of the Fourier you only have two photoreceptor types I mean that's an anatomical fact you can show that psychophysically that it is true that in the central part you only have access to red and quote red and quote green photoreceptor signals yet again I don't see that you know if I look at the blue if I look at something blue like the screen here it doesn't look like when I really focus at the center that I don't that there's a hole there that something looks yellow then you have the blind spot where you don't have any information direct information then you constantly move your eyes and if you do if you just try to mimic this as you'll see in the movie you know you get this horrible blurred signal and furthermore something I have mentioned all yet you also blink you blink like not quite as often as you move your eyes you blink two or three times a second Pali to lubricate of course you can also do it voluntarily that has a different dynamic so Pali you do that to keep the eye wet and lubricate Pali you know when you have dust when there's dust or something to clear that Pali it also seems to reflect a certain autonomic state I mean when people are very nervous something they might also blink but again that doesn't really make until I put this out to you I'm not sure when the last time you thought when you actually constantly knew you made a blink probably many years ago you do two or three times a minute so let me show you this movie so it's a simulation it's now a few years old so first he scanned using a different signal Laurent scanned so this is what a person would do looking at the looking at the at this image so the point of sharpest thing is always here the blind spot is always here here the point of sharpest thing that's why you see this funny color you have no blue sampling and of course it falls off so that's how the output of this image would be mimicking something about the this is in real time now you'll agree with me that vision doesn't really look like that now of course it's a computer simulation and all of that but I mean it's trying to get this I think it just loops now and here you're sitting actively on the car here you're sitting on the optic nerve as it were that's why with respect to you we can look at once more with respect to you the blind spot doesn't change and the point at the center where you don't have blue photoreceptors doesn't change and of course you see the complement of blue which is yellow this is what the external observer views and now in a second you'll see what the eye would see this is first I mean at slow speed of course here we didn't bother to model the edge of the visual field that's why you see the edges okay and let's finish with that so this is just to underlie the message you don't see with your eyes