Faceted eyes - pros and cons

In general, in their (perhaps fric?) In this reasoning, I decided to first prove a small theorem.

Consider an ordinary simple eye, with the lens cornea and vitreous body transparent to infracassal radiation. Let's put this eye in a thermos. The temperature of the eye, the air in the thermos, the retina and the walls of the thermos are the same. The system is in thermal equilibrium. At the same time, the eye looks at the picture on the wall of the thermos. Infrared radiation forms an image on the retina. At the same time, since infrared radiation hits the retina, it brings some amount of energy. But the retina does not heat up, because all the excess heat is carried away in various ways. For example, the retina itself emits infrared radiation, which falls on the picture on the wall of a thermos flask. In general, the laws of physics prohibit such a system from changing the temperature in any way due to heat transfer alone, since the system is already in thermal equilibrium.

Now, suppose that it is possible to create such an optical device consisting of a set of lenses that increases the brightness of the image on the retina. If this is possible, the retina will receive more energy over the same area than in the previous example. At the same time, its heat loss will not change - the retina's own radiation depends on the temperature. This means that the retina will heat up above the equilibrium temperature to radiate excess heat. And this will lead to the fact that such an eye in a thermos can be made into a perpetual motion machine - the retina can be used as a heater, and the wall of the thermos can be used as a refrigerator. This means that our assumption was wrong, and it is impossible to create such an optical device that increases the brightness of the image.

(then I'll add a picture.

The only way to increase the brightness of the image on the retina is to increase the size of the pupil. In this case, the increase in brightness is compensated by an increase in radiation in the opposite direction through the dilated pupil. In this case, the brightness can be increased only if the pupil is enlarged, without equalizing the distance between the retina and the lens. If you simply increase the size of the eye, then the additional radiation will be distributed to the enlarged image, and there will be no gain in brightness. Is it possible to make the pupil larger than the distance between the lens and the retina? This is not directly prohibited, but such an eye will have a very inconvenient design and will not be suitable for living organisms.

But then a strange question arises:
Why do many nocturnal animals have very large eyes? After all, simply increasing the size does not give a gain in brightness? And most importantly, why is it that a pair of binoculars, spyglass, or telescope can see poorly lit objects better than the naked eye?

I will assume that it's all about the principles of how the retina works.
Why can't a person see in low light? In a room where almost nothing is visible anymore, quite a lot of visible light photons still fly into the gas. What prevents you from catching them? Lack of retinal sensitivity? But, in theory, no matter how weak the signal is, amplifying it seems to be a trivial task. Couldn't evolution have chosen a different synapse structure for nocturnal animals, but instead started building huge eyes? I will assume that the whole thing is in the internal noise of the retina. The sensitive elements of the eye not only pick up a useful signal, but also create some noise level that mixes with the useful signal. And if the output signal of the cone is simply amplified, then nothing but noise will be visible. Nerve cells in the retina are already pre-processing the signal so that no noise enters the brain. But in low light, they also cut off the useful signal.

How can you see a little more in the dark? In theory, noise is a random process. This means that if you take the arithmetic mean of several neighboring cones, then the noise will subtract itself, as it were, and the useful signal will increase. It is this processing that allows you to see at least large objects in low light. But at the same time, the sharpness of vision decreases - since it is necessary to add up the signal from neighboring cones.

The noise hypothesis also explains another effect: why frightened or excited animals have dilated pupils. One of the ways to deal with noise can be to repeatedly add the signal to itself from the same cone, obtained at a previous time. This allows you to get a clearer image, but the addition process itself is time-consuming. When there is no hurry and there is enough light, the pupil is narrowed. In case of danger, the pupil expands to reduce the reaction time due to fewer iterations of addition.


Visual acuity is also affected by other factors.
The imperfection of the eye shape itself leads to various types of aberrations, due to which the image is blurred.
With an ideal optical system, the wave nature of light becomes noticeable - the light bends around the edges of the pupil and makes the image more blurred.
Increasing the size of the eye solves both of these problems.

In general, the conclusions are as follows: with increasing eye size, most of its characteristics improve.:
1. The ability to see in low light increases, since the image is larger in the large eye, and you can add up the signal from more sensitive elements, with less loss of visual acuity.
2. Reaction speed. Larger image - you can add signals from neighboring elements instead of long-term addition of signals from one sensitive element. It's faster.
3. Visual acuity. The larger the image, the easier it is to see. This is obvious.
4. Imperfections in the optical system and problems caused by the wave nature of light are also treated by increasing the size.

In general, the bigger the eyes, the better it sees.

And then the question arises, why do insects use their complex eyes instead of 1-4 simple eyes? After all, each element of the complex insect eye sees very poorly, for the above reasons. At the same time, in order to see something, complex eyes have to be made huge. Wouldn't it have been easier to make simple eyes?

The faceted eye is not fundamentally different from the" simple eye "of vertebrates, except for the presence of" reinforcement " in the form of facets.
As far as I know, in insects, each ommatidium is an isolated optical system... And that, I think, is what is most hindering him. Small eye diffraction from the edges of the ommatidium should not interfere so much.

Diffraction does not interfere there, each facet is a separate "pixel", although it is quite complex.

> And there does not interfere with diffraction, each facet is a separate "pixel", although it is quite complex.
this should be the most annoying thing to do. due to the fact that each facet is small, there is a noticeable diffraction. Which causes the facet to receive light not only from the direction it is directed, but also from neighboring directions. At the same time, each facet is a separate pixel, which means that signals from neighboring pixels will be strongly mixed as a result of diffraction and it will be difficult to build a clear picture from them.

In general, diffraction is the most obvious, but not the only problem with faceted eyes.

I, of course, am far from physics, but something tells me that for such an effect of "steering light around an angle", the width of the ommatid should be comparable to the wavelength of light, and in nature it is very much larger. As for the darkness, almost all crepuscular and nocturnal insects that actively use vision, the walls of the ommatidium become transparent at night and turn out to be something like an ordinary eye with a large pupil, although a very clumsy "glass-like body".

not so "very much more". let's say the wavelength of some green color is 0.5 microns, and the size of ommatidium is 5-50 microns. the difference is 10-100 times. in general, the extent to which the light comes around the corner can be assessed by the Ray criterion. two light sources can be distinguished from each other when
d/D>l/A, where d is the distance between the sources, D is the distance from the sources to the eye, l is the wavelength,and A is the pupil diameter.

let's say an insect looks at two green leaves, 5 cm apart and 1 meter away from the insect. the wavelength is 0.5 microns, the eye size is 10 microns. we substitute them into the formula, and we see that the insect will perceive them as one leaf. accordingly, a predator of 5 cm in size can wave its legs as it pleases, and the insect will perceive it as a stationary object.

at the same time, only 100 ommatidiums in width are placed in the eye with a diameter of 1 mm.

now let's compare it with a simple eye. we make a simple eye with a size of 1 mm and a pupil of 250 microns. now from one meter you can distinguish objects with sizes from 3 mm.

This post was edited by Silly Fairy - 11.08.2020 08: 11

in the compound eye, each ommatidium looks to its own sector. This means that light arrives only from its own sector. Although I have already shown that increasing the size of the eye does not give a gain in illumination, a complex eye does not have a very favorable ratio between signal and noise. due to the fact that the ommatidium itself is very small, it also forms a very small image. and it's hard to fit enough sensitive cells into this small image. and placing sensitive cells out of focus will either further reduce the brightness, or increase the effect of "getting light from around the corner. this impairs two of the most important characteristics of the eye-the speed of reaction and the ability to see in low light.

The ommatidium alone does not form any image at all. The facet width in well-seeing insects is significantly larger than 10 microns. And I do not know such an eye with a width of 1 mm, where as many as 100 ommatidii would fit in one row. The general rule is that the smaller the eye, the fewer facets it has. About distinguishing the movement of a predator, you are very wrong. Even if the insect does not see its details and exact dimensions, the movements will affect the change in light in this area, this is enough. In general, well-seeing insects perceive movements, especially fast ones, much better than you and me, their "frame rate" is much higher. To realize this, just try to catch a fly with your hand. But the insect does not need to count the leaves, it is enough to know that they are there.

And yet it is not clear what the problem is? It seems obvious that well-developed vertebrate eyes can see better than well-developed insect eyes. So what? )

And insects have no problems with the horseshoes, as well as insects-predators with their victims)
Just recently I saw such a straight Hollywood story, I stand, watch, and suddenly I see a tarantula literally flying over the grass! He does not just run, but jumps like a horse at a gallop, pushing off from the blades of grass! Then the cryptoheilus flies out to him from somewhere, and the chase begins! This is something, the poor tarantula jumps from side to side, the wasp then lags behind him by a couple of tens of centimeters, then overtakes almost, and in the end the wasp caught him! so, all this, well, as it seemed to me, lasted less than a second, and the tarantula ran clearly more than a couple of meters!
Apparently they missed each other in the tarantula's burrow, and the tarantula was able to jump out and tried to escape, he had a head start until the wasp also got out, and all the same - it didn't help, noticed, and overtook
Something seems to me, with the vision of our wasp everything is fine, as well as in other insects)

But yes, if the tarantula was not so stupid, and hid after a little ran away from the mink, then the wasp would not have found it. Well, maybe just by accident. But the tarantula is stupid, does not know about the peculiarities of insect vision, and runs, thereby betraying itself)

I could have found it easily by smell. Not with a single vision. So the spider did the right thing, but it was out of luck. Here in the deserts there is his relative, who rolls from the wasp down the slope of a dune and usually rolls away safely.

What does pan mean by "seeing better"? The" well-developed eyes of vertebrates "have only one distinct advantage over the" well-developed eyes of insects". But this is not the only parameter of "good vision". For example, insects are undoubtedly in the lead in terms of "rate of fire", and mantis crayfish is in the lead in terms of color perception. About photosensitivity-H. Z., it is necessary to check.

The message was edited INO-11.08.2020 15: 29

I do not dispute that everything is OK with insect vision, but it is still unclear why they use faceted eyes, when at first glance it seems that simple eyes are better in all characteristics, including those that are critical for insects.


>The fact that well-developed vertebrate eyes see better than well-developed insect eyes, it seems to be obvious. So what? )

It is not clear why insects use faceted eyes when simple ones are much better.

>One ommatidium does not form any image at all
It seems that, as far as I understand the structure of the ommatidium, it quite forms. There's a cornea and a crystal cone that form the image. In the center of the image is a rhabdom that captures the central pixel of the image. The other part of the image is scattered and absorbed by the pigment cells - the insect does not see it. But that doesn't mean it doesn't exist. In particular, this image consumes light energy, which is captured by the eye. A situation where most of the energy that has already passed through the optical system of the eye and formed an image is uselessly lost already looks wasteful.

> Even if the insect does not see its details and exact dimensions, the movements will affect the change in light in this area, this is enough.
Experimental observations of radars, cameras, eyes, antennas, and hydrophones show a simple pattern : the greater the total power that the receiver receives, the better the target is visible. And a large antenna / pupil / lens certainly gives an advantage.

Not always a large antenna / pupil/lens gives an advantage in energy flux density, but it seems that increasing the image size (and with it, the power) is useful.

Complex eyes are wasteful in terms of size. In a simple eye, all rays coming from a point target and passing through the pupil fall into the image of this target on the retina. If the insect is looking at a pinpoint target, it sees it with only one ommatidium. And the retina captures only the radiation that has passed through the small pupil of this ommatidium. Of course, the power will be much less than that of a simple eye of the same size. And the reaction time is worse, and the ability to see in low light, and the ability to see the target in general.

For example, a simple eye looking at a red berry against a background of green leaves can already instantly find it by its characteristic color. For a compound eye of the same size, the image of a berry will blend in with the image of the surrounding leaves and it will be much more difficult to see it. To see as well as with a simple eye, you have to do more.

One pixel is not an image, is it?

> I don't know what mathematics says about this, but nature knows better.
Naturally, nature knows better, but I want to understand why nature made this decision. However, nature can also be wrong, we have already given an example of the human eye, in which light first passes through a layer of nerve cells, and then hits the sensory cells.

(you will laugh, but for some types of photo/video camera matrices, the conductor layer also blocks the sensitive elements, and is also located through the ass (but the matrix can be made from one layer - cheaper). So humanity has copied nature's obvious flaw. )

> The large pupil has the opposite side in the form of a small GRP, and therefore, the need for constant accommodation. And while the eye focuses on a fast-moving object to determine whether it is a friend or an enemy, its owner may already be caught.

The objection looks strong. Earlier, I also thought that this was the advantage of faceted eyes. But then there were doubts about the strength of this argument. First you need to figure out what kind of FLU insects need?
In theory, an enemy the size of an insect can be seen by a simple eye at any distance at which there is enough sensitivity - in the distance it sees it clearly due to the setting "to infinity", near - not clearly, but since the enemy has large angular dimensions near, this is no longer a problem.


They also see polarized light.
This is the advantage of naseocids over vertebrates, but not of faceted eyes over simple ones. Nothing prevents you from putting the same sensitive elements in a simple eye as in a faceted one.

> However, in horse-riding spiders, simple eyes are not less effective (but rather more), but they are not inverted, and, apparently, they really capture light very effectively.
This is a very clear example of an effective simple eye. Something is starting to feel like spiders have the best eyesight in the animal kingdom. Yes, some eagle has a higher resolution, but the spider can grow to the size of an eagle... And he would have learned to fly...

IMHO, spiders have the most convenient number of eyes: they do not have the problems of faceted vision described above, they have small eyes for looking close and large ones for looking into the distance. By adding up images of different eyes, you can get a circular view, and by adding up the same parts of the image, you can increase the brightness and reaction speed.

As for skakunchikov, I Googled something here and it seems that so far I have found that skakunchikov has up to 1000 receptors in the retina of large eyes, that's all! + another thing is that skaters have color vision, and probably not all the information goes to full resolution. Yes, there they are not inverted, and the environment is very transparent, but still, and at the same time, the angle of the "objective" of skakunchikov is very narrow.
Some bee or wasp with such eyes probably would not be able to find a flower at all or get into a bee's nest only in one eye there are about 5000 ommatidiums, in which there are 9 nerve cells and several dozen of their processes.
And the dragonfly??

To distinguish the polarization, the surface of ommatidium, which has the properties of a polarizer, is somehow cunningly fragmented, and in different facets the visual cells are located differently, with a shift under this fragmentation, according to this scheme, it seems impossible to repeat this in the ordinary eye. Maybe some other scheme is possible, but alas-apparently, this was not done according to any scheme, and it is not necessary! Having eyes with much higher resolution, you can do without any polarization

About the alleged "irrationality" of vertebrate eyes in the sense that nerve fibers go on top of the receptors themselves, which "militant atheists" like to use very much as an argument against "intelligent creation".
In fact, there is a lot of sense in this, which consists, like, in the contact of the receptor bodies with the vascular network, which improves the functioning of the eye

In general, all these attempts to find "errors of nature" are actually, to put it mildly, not serious, for this you need to understand at a very high level everything that you are trying to criticize, but so high that you can create something like this yourself.
This is similar to attempts to criticize the device of modern cameras, for example, that it would be much better to make the matrix spherical, they say, okay, before the film did not allow this, but now that.
But in fact, this is such a pain in the ass that the advantage of a flatter image field pales before other problems.

Polarization is clearly needed by bees not only as a tool for collective orientation, but also simply to compensate for the imperfection of eye resolution in remembering the path. With better vision and more developed communication, this data can be transmitted and remembered much more accurately based on visual landmarks, which is what people actually do
And I have a" Viking stone", with the help of which you can find the exact location of the sun when it has set below the horizon, while there is still at least some twilight, but nothing more, this stone can not replace the compass

And if you turn it the other way around, the receptors will be further away from the vascular network, but I didn't find anything about it, but I found this
https://www.gazeta.ru/science/news/2015/03/...n_6970797.shtml

And what could be more than that? Determining the position of the sun at night or something?
The essence of the stone was to find the sun in a cloudy sky, this is clearly stated in the sagas and chronicles. Local landmarks are good, and stingy hymenoptera can use them as well as we can - but even a human can't easily explain the route to another person using these landmarks. And bees can explain by the sun.

>In general, all these attempts to find "errors of nature" are actually to put it mildly
, which is why I took the nickname "stupid fairy"
at the expense of polarization... In any case, it is read through a polarizing filter of one or another design. And I still don't see what prevents me from picking up the polarization with a simple eye.

and yes, recognizing polarization is a useful skill. since to find the way in other ways, you need more brains, with which insects are strained...

This filter is the faceted eye itself. On a simple eye, it would be necessary to build up a tricky polarizing film, which nature did not think of.

Meanwhile, the key parameters of eye optics are quite easy to calculate. I suggest you do a little research: Someone will name the specific vision parameters of an insect, and I will try to tell you what the parameters will be for a simple eye that gives the same image.
Interested in:
- Size of the eye, mm
.- Number of ommatidiums (pieces)
- Field of view (vertical and horizontal, in degrees. approximately).

But first we need to answer one more question... And what are the parameters of the FLU required by insects? And what typical objects do they need to see?
In theory, the maximum radius of the grip is easiest to make infinite, especially since the "infinity" can start as early as 20 cm. A minimal one... How long are the antennae? so that what the insect can no longer see can be felt... But this is not a mandatory requirement. For example, if a hypothetical insect with simple eyes, an insect suffering from farsightedness, can see the legs of its relative up to zero distance, will it be enough?

Only a spherical horse is easily calculated in a vacuum, even the lens of their non-so many fixed spherical glass pieces will not be easy to calculate. And the eye is completely aspherical, and in the case of vertebrates, it is adaptive. Plus very specific parameters of the "matrix".

Or do you mean by "key parameters of optics" only the GRIP?

Obviously, you should ask insects what kind of GRP they need. But even without calculations, it is obvious that in the appositional eye it is huge.

So we don't need precise and particularly complete calculations. The accuracy of "Plus or minus bast shoes" is quite enough. You don't even need to count the specific parameters of the lens and roughly estimate what resolution it can give, what will be the GRIP, sensitivity, etc.

If we are trying to figure out which eye sees best in reality, bast shoes will not help. The actual resolution of the lens is difficult to calculate, and for the eye it is hardly possible at all. The curvature of lenses (and there is also aspherics everywhere), refractive indices and light transmission - you don't know all this, and neither do I. And you will not determine the photosensitivity of the receptors by any formula at all, only experimentally. So you can only count a spherical horse out of touch with reality, alas.

I suggest at least counting a spherical insect in a vacuum.

What for? After all, they do not exist in nature.

I recently observed my pet praying mantises - I saw another advantage of the faceted eye - it has the same image quality for any angle of incidence of rays (within the angle of view). No need to turn your head or move your eyes. But. as always, permission was brought to the Women's Council: one small object can only see a small part of the video material at a time, which is already not very thick there. But there is a cool effect of "Giaconda's gaze" - from which side you can't look into the eye of the praying mantis, the "pupil" looks exactly at you. But this is in the daytime mode, at night everything is more interesting.