2D becomes 3D when looked at by one eye

[M Grandin]

When I was about 25 I noticed that a 2D photo appeared 3D when looked at through a magnifying glass. First I thought it depended on the lens, but then realized it was because I looked at the picture by just one eye - exactly as the camera lens had seen it in reality. I.e. there is much 3D information in the picture, but discarded as 3D when both eyes simultaneously (unconsciously) judge it as 2D. I showed this to some friends and almost all of them agreed
I was right. Most of them also thought my explanation for this "phenomenon" was reasonable. Although it didn´t work for everyone. But it is not a faint farfetched illusion - rather surprising and to some people even scaring.

But first after a moment of perhaps 10 - 20 sec the 2D picture "jumps" into 3D.

 

[ArtistIC]

I will give a basic idea of an explanation. What you perceive as 3d is what your eyes and brain interpret as they combine information from your left and right eye. If you close one eye you immediately lose the pereception of distance and only guess how far an object is by knowing how far an object is judging by its size. Now if you look at an object (like a 2d photo) with one eye through a lense and and the other as usual you get a set of distorted information to your brain that "trick" it in believing there is actual depth in your photo.

Same trick is applied in cinema 3d glasses where one lense is darker than the other tricking the eye by causing a delay to process of one eye (darker colour) hence givingh the false impresion of depth as images viewed are not matched.

My reply is very rough but would give you a first idea before someone jumps in and gives a more scientific explanation :)

 

[MikeGomez]

I have to admit, when I first read this post I thought what you were talking about was silly. Good thing I tried it before responding, because it seems you are correct. Pretty interesting!

There are several cues for 3D vision which come to mind. Each eye sees object at a different angle. This effect is greater for closer objects than for distant ones. Each eye sees objects at an offset position, This is the parallax effect. There are lighting and shadow cues. There are size cues, and other cues such as the texture of the materials which you are viewing.

I think your analysis seems fairly reasonable.

there is much 3D information in the picture, but discarded as 3D when both eyes simultaneously (unconsciously) judge it as 2D.

One eye viewing something, even a 2D picture without angle and parallax information, still has certain 3D lighting and sizing cues. Our brains still processes a lot of 3D information. Therefore viewing a 2D image with one eye, we still seem to get a reasonable sense of 3D.

However, when both eyes are presented with a single 2D picture, the missing 3D cues tell us that we are viewing only a 2D image. It seems like an unexpected result, but true nonetheless. Nice job.

 

[M Grandin]

I have to admit, when I first read this post I thought what you were talking about was silly. Good thing I tried it before responding, because it seems you are correct. Pretty interesting!

There are several cues for 3D vision which come to mind. Each eye sees object at a different angle. This effect is greater for closer objects than for distant ones. Each eye sees objects at an offset position, This is the parallax effect. There are lighting and shadow cues. There are size cues, and other cues such as the texture of the materials which you are viewing.

I think your analysis seems fairly reasonable.




One eye viewing something, even a 2D picture without angle and parallax information, still has certain 3D lighting and sizing cues. Our brains still processes a lot of 3D information. Therefore viewing a 2D image with one eye, we still seem to get a reasonable sense of 3D.

However, when both eyes are presented with a single 2D picture, the missing 3D cues tell us that we are viewing only a 2D image. It seems like an unexpected result, but true nonetheless. Nice job.

Thanks for kind words !

I think this "phenomenon" should be wellknown, but none appears having heard of it.

 

[sophiecentaur]

Thanks for kind words !

I think this "phenomenon" should be well known, but none appears having heard of it.

This "phenomenon" has been well know by painters for centuries, I think. (Consider the really poor depth in earlier, medieval church paintings, done before the 'tricks' had been learned.) Your brain takes clues where it can find them. If it sees a printed picture on a flat sheet then it will realize it is 2D, due to binocular vision. To give a good impression of depth to someone who looks with both eyes you need some greater impact to overcome the strong clue that it's painted on a flat sheet. Vivid colours in a foreground and more desaturated colours in distant parts of a scene ('the perspective of Light' is described by Leonardo Davinci). There are dozens of optical illusions that give unnaturally vivid depth effects.
With a very 'ordinary' photograph, those clues will not necessarily be there and the 2Dness of the picture will beat the 3Dness in your brain. I think that's one of the things about ones own 'successful' pictures - you were just lucky at the time.

 

[M Grandin]

I will give a basic idea of an explanation. What you perceive as 3d is what your eyes and brain interpret as they combine information from your left and right eye. If you close one eye you immediately lose the pereception of distance and only guess how far an object is by knowing how far an object is judging by its size. Now if you look at an object (like a 2d photo) with one eye through a lense and and the other as usual you get a set of distorted information to your brain that "trick" it in believing there is actual depth in your photo.

Same trick is applied in cinema 3d glasses where one lense is darker than the other tricking the eye by causing a delay to process of one eye (darker colour) hence givingh the false impresion of depth as images viewed are not matched.

My reply is very rough but would give you a first idea before someone jumps in and gives a more scientific explanation :)

Are you giving me a "first clue" to this "phenomenon"? I had already explained it in original post.
Your closed or "absent" eye does not contribute to the illusion - in other way than not giving information .

You maintain 3D movies use this phenomenon. I am not upgraded regarding 3D movies/videos - but presume these, in some way or other, use separate pictures for left and right eye respectively. By using swift electronics in glasses or some kind of raster-technique etc. Cameras for 3D pictures/movies have two lenses.

 

[M Grandin]

This "phenomenon" has been well know by painters for centuries, I think. (Consider the really poor depth in earlier, medieval church paintings, done before the 'tricks' had been learned.) Your brain takes clues where it can find them. If it sees a printed picture on a flat sheet then it will realize it is 2D, due to binocular vision. To give a good impression of depth to someone who looks with both eyes you need some greater impact to overcome the strong clue that it's painted on a flat sheet. Vivid colours in a foreground and more desaturated colours in distant parts of a scene ('the perspective of Light' is described by Leonardo Davinci). There are dozens of optical illusions that give unnaturally vivid depth effects.
With a very 'ordinary' photograph, those clues will not necessarily be there and the 2Dness of the picture will beat the 3Dness in your brain. I think that's one of the things about ones own 'successful' pictures - you were just lucky at the time.

You have obviously misunderstood what I wrote. I talk about a mental mechanism that translates a 2D picture or photo into apparent true 3D, when looked at by one eye. This phenomenon obviously works regarding ANY picture or photo - but in different extent depending on circumstances. The effect takes place after appr 1 - 10 seconds, not after 10 -20 seconds, as I for some reaoson wrote in original post.

This is obviously not a known phenomenon - in fact I have never heard of anyone having heard of it before I mentioned it. Some people even deny the possibility of this and object against trying it.

Painters giving 3D-impression on their pictures has nothing to do with this. Although the effect
also works on paintings: Looked at by one eye the painting looks more like true 3D.

 

[rcgldr]

Same trick is applied in cinema 3d glasses where one lense is darker than the other.

Modern cinema 3d glasses are polarized, and two polarized images are shown on a common screen. The angle between the polarized glasses and images are 90°, ideally completely blocking out the "non-aligned" image going throug a lens.

Old cinema 3d glasses were red + blue, and two images biased towards red and blue were shown on the screen.

Some home 3d glasses are shuttered and block / unblock vision in each eye, in sync with alternating images shown on a screen / monitor.

Lenticular printing is another way to create a 3d image.

 

[sophiecentaur]

Modern cinema 3d glasses are polarized, and two polarized images are shown on a common screen. The angle between the polarized glasses and images are 90°, ideally completely blocking out the "non-aligned" image going throug a lens.

Old cinema 3d glasses were red + blue, and two images biased towards red and blue were shown on the screen.

Some home 3d glasses are shuttered and block / unblock vision in each eye, in sync with alternating images shown on a screen / monitor.

Lenticular printing is another way to create a 3d image.

Stereoscopic Cinema doesn't use plane polarised light - if it did, tilting your head a small amount would totally ruin the effect because you would get both pictures in both eyes. They use Circular polarisation clockwise and anticlockwise for the L and R eyes.

Stereo cinema is very impressive but it is actually pretty over-stated, in order to sell that particular effect. Using the two cameras, separated a bit more than human eyes, will exaggerate the stereo effect. This is easily demonstrated with a large pair of binoculars, where the objective lenses are much further apart than your eyes. There can be an unsettling level of stereo effect, which heightens the impact of what you see. The super-3D effects that place images righ on the end of your nose are quite tiresome, after a while and suited mostly to animations.
The other, more subtle distance clues which good pictures contain will give good depth perception at distances well beyond those that an eye separation of 70mm can resolve stereoscopically. Our stereo vision (as potential hunters) is at its best at the sort of distances where leaping on and catching prey is helped by our vision - i.e. up to, perhaps 10m. Beyond this, we use parallax, colour saturation and perspective clues at least as much.

I have a feeling that stereo will become much more subtle in cinema, in future. In the early days of colour TV, people used to crank up the colour saturation control - just to prove to themselves they actually had a colour set! The same thing will happen with stereo tv and cinema, which is ridiculously over-done these days. Technology will have to advance a long way, however, and to do away with special spectacles and the like, before it will be really worth the bother of watching.

 

[plasmon]

Monocular vision gives only the quality of depth and it needs lots of variation in texture, shading and a strong perspective(not available ideally in photo, possible in painting) to construct a 3d scene. In contrast, binocular vision gives a precise estimate of depth. It only needs a minimum of two objects(without texture or shade) to quickly construct a basic 3d scene. Monocular cues complement stereo but can never really compensate for the lack of latter.

 

[sophiecentaur]

I think you are overrating the importance of binocular vision. It is a very convenient time saver over a useful but small range of distances. (Suits close up hunting skills of course.)
A huge proportion of the population has vision in one of their eyes that's limited in some way. Our internal map of the local world is assembled from many clues, not all are even visual. Machine vision may well use exclusive binocular vision but our brains are far better.
I know two people who do not use binocular vision due to strabismus when they were young. Both can play raquet games fine.

 

[Delta Kilo]

Next time try it while playing a decent 3D computer game - it will blow your mind!

I think this "phenomenon" should be wellknown, but none appears having heard of it.

Well, it is actually well-known. It is associated with paintings mostly - when looking at one you are supposed to lean back with your arms crossed and one eye closed :) I guess paintings are involved because for centuries these were the only 2D reproductions of 3D scene available.

Stereo TV has its own share of problems. Being "more physically correct" it actually limits director's creativity. For example, we are quite used to abrupt changes of scale in ordinary 2D movies. One minute there is a tiny little person on a big stage and the next moment there is one big head filling the entire screen. It looks all right in 2D but in 3D it's just hard to shake the feeling that you are looking at a little animated doll, and then suddenly you are looking at a head 3 times bigger than normal. It fails because 3D TV is limited in the range of depths it can render and confusing stereo cues make it worse than no cues at all.

 

[sophiecentaur]

Next time try it while playing a decent 3D computer game - it will blow your mind!Well, it is actually well-known. It is associated with paintings mostly - when looking at one you are supposed to lean back with your arms crossed and one eye closed :) I guess paintings are involved because for centuries these were the only 2D reproductions of 3D scene available.

Stereo TV has its own share of problems. Being "more physically correct" it actually limits director's creativity. For example, we are quite used to abrupt changes of scale in ordinary 2D movies. One minute there is a tiny little person on a big stage and the next moment there is one big head filling the entire screen. It looks all right in 2D but in 3D it's just hard to shake the feeling that you are looking at a little animated doll, and then suddenly you are looking at a head 3 times bigger than normal. It fails because 3D TV is limited in the range of depths it can render and confusing stereo cues make it worse than no cues at all.

The fixed camera position is a major limitation. Even with a stereo system the two cameras are fixed. One eye and a head with controlled movement is more than adequate. Stereo is full of possibilities for optical illusions.

 

[plasmon]

For centuries, we are using beauty to see space but with stereoscopic tools, it is time to use space to display beauty.

 

[Suresh1971]

I had received a video which when viewed with one eye had fantastic visual depth! I couldn't believe that it's possible to view a 2d image as 3d with just one eye! Was researching on Google and found that a scientific paper was published recently and also came across your site.

Other people who are sceptic or who don't believe it please check the embedded video on YouTube in full screen.

 

[sophiecentaur]

I had received a video which when viewed with one eye had fantastic visual depth! I couldn't believe that it's possible to view a 2d image as 3d with just one eye! Was researching on Google and found that a scientific paper was published recently and also came across your site.

Other people who are sceptic or who don't believe it please check the embedded video on YouTube in full screen.

The distinction between binocular and monocular vision is very overstated in home brewed models of the way vision works. It seems to be assumed that the two images on our retinas are what counts in our appreciation of space. It it were really like that, one eyed people and most herbivores (with 360° visual field) would have no depth perception at all. Our 'view' of the space around us is not only provided by binocular vision. Many blind people have as good an idea of the layout of a room as a sighted person's mental image of the bits of a room that they are not actually looking at. Of course we make use of binocular vision when playing games or hunting but for stationary scenes, the depth information is just as available, once we start to move about. A baseline of five or six cm only helps significantly within 'pouncing distance' or in ball games, which are not the major parts of most of our lives.
Stereo programme material can be stunning and entertaining but, the low takeup of stereo TV tells the real story; Stereoscopy contributes very little to our experience of the world. Big and sharp beats 3D TV any day because it's what our brains crave for.

 

[JT Smith]

Pretty cool phenomenon. I hadn't heard of it before but it jumped right out at me when I tried it.

Some interesting experiments here: https://www.ncbi.nlm.nih.gov/pubmed/23892232

I tried using an aperture and didn't notice that it made very much of a difference for me. Maybe some other details of their procedure are important.

They point out that the experience of 3D vision and the perception of depth are not one in the same. That is, you can perceive depth in a 2D photo with both eyes open; you just don't "feel" it.

 

[DaveC426913]

I'd like to experience this.
I tried watching the apple video posted by sc with one eye closed. No unusual effect.
What am I supposed to experience? Should it seem like the apples are literally floating in front of my screen?

 

[JT Smith]

Not floating in the front of the screen, but a pronounced effect; one that you feel as much as see. In that study I linked the participants expressed surprise when they experienced it. It's hard to describe but you'd know it if you saw it.

It's worth noting that a small fraction of the people in that study didn't have the experience.

 

[sophiecentaur]

the experience of 3D vision

We do not have "3D VISION". We have binocular vision. True 3D vision would allow us to see what's behind objects in front of us. What ie do see, in fact, is a small part of the background that's seen by one eye but not the other. There could be absolutely anything behind an object but we make assumptions which can really fool us into believing what we have actually seen.
In that video sequence of apple trees, we are shown all of what's behind the foreground trees because the camera tracks over a long distance - much more than the distance between eyes. Our brain (clever little devil) remembers all the images and brings them together to get a 'true' 3D model of what's there. Needless to say it is not a 'true' model but it's much better than the inadequate model that our two eyes give us. Our binocular version is quite adequate for what we need most of the time.

 

[Bystander]

See "holography, perspective, focus, depth of field, animation."

 

[sophiecentaur]

See "holography,

Holography is not stereoscopy it can produce images (that have limited resolution ) that will show the three dimensional information of an object. Some visual holograms can be rotated to show good views of a scene from all angles, But even a hologram doesn't show the hidden internal details of the rooms of a house. There is a limit to the amount of 3D information even in a hologram.

 

[JT Smith]

We do not have "3D VISION". We have binocular vision. True 3D vision would allow us to see what's behind objects in front of us.

Okay, fine. Semantic nitpick accepted.

The conclusion by those researches was that "stereopsis is not a simple by-product of binocular vision or visual parallax", but "can occur for static two-dimensional pictures without binocular vision".

 

[sophiecentaur]

Okay, fine. Semantic nitpick accepted.

It is not a "nitpick". We have binocular vision and that is what it is. It is not "3D" vision. We make the best 3D model in our brains from the binocular images we get (or without them is they are not available). There are many illusions that make use of this and which make good cinema. I have no problem with them. My problem is with the over simplistic way in which binocular vision is used as a reason for our 3D perception. As with colour vision and sound perception, the popular description falls far short of an understanding - or even an appreciation of the sort of complex processing that our brain is doing with the information it receives.
Example: The eye's lens forms an inverted image on the retina. This is, somehow a big deal and the brain, apparently has its work cut out to deal with this inversion - interest then fizzles out. I am not claiming that any contributors to this thread are guilty of anything so superficial but this sort of thread always has a risk of heading in that sort of direction.

 

[JT Smith]

It is not a "nitpick". We have binocular vision and that is what it is. It is not "3D" vision.

I don't know what the precise definition of "3D vision" is, assuming there is an agreement about that. But in this context, where the authors of the paper I was citing actually use the phrase "seeing in 3-D", I thought it was clear enough what I meant.

I love to nitpick too. But I recognize it as a vice.

 

[sophiecentaur]

I don't know what the precise definition of "3D vision" is, assuming there is an agreement about that. But in this context, where the authors of the paper I was citing actually use the phrase "seeing in 3-D", I thought it was clear enough what I meant.

I love to nitpick too. But I recognize it as a vice.

My problem is that there is an implication that, viewing with one eye should somehow prevent you from or detract from getting 3D information of a scene. It is a serious over simplification of the process. That video demonstrates to me that binocular vision is not necessarily a help in 'seeing' a 3D scene. The title of the thread implies that this is a big surprise but why should it be?

In that video (which is on a screen, say 60cm away from you) your binocular vision is telling you that the objects you are seeing are in a single plane (which they are, actually). When you look with one eye, you lose that information about the distance from the screen (your head is stationary so no parallax help) and start to use the other, in this case very strong, clues about the scene. If you were actually viewing that scene directly in place of the camera, you would be getting lots of information about depth from the non stereoscopic information. Stereoscopic vision could be telling you that there is some depth to the scene ( that not everything is just 60cm away, as on a TV screen) but it's all the other clues that go together to form the internal model of the scene. Can you really argue that this paragraph is a 'nitpick'?

EDIT: PS @DaveC426913 if you read what I wrote and look at the video again, you could, perhaps convince yourself about the effect. It might be that one of your eyes is very dominant and that you do not get strong stereo information. My Daughter in Law grew up with strabismus and, even though it has been largely corrected, never learned to use stereo vision effectively (so she tells me). She can choose to look at an object with one eye or the other, at will but can't actually get the 'natural' stereo effect that most people do. Perhaps if you look at the edge of the screen. the plane of the picture may get into your brain better and you might see a 'flatter' scene' I see the effect very clearly and it confirms my idea about the reason for it.

 

[DaveC426913]

EDIT: PS @DaveC426913 if you read what I wrote and look at the video again, you could, perhaps convince yourself about the effect. It might be that one of your eyes is very dominant

This is actually quite true. While both my eyes are equally adept at seeing (I don't have any kind of partial vision loss), I have always habitually read with one eye firmly closed, so I look like a pirate missing his patch. My right eye is very dominant over my left.

TMI: I attribute it more to a difficulty keeping them both pointed at the same spot. When fatigued, it is very difficult to keep my eyes from going "de-crossed" (wall-eyed). And they rotate. If I stop trying, the image from one eye might easily rotate 20 degrees from the other eye.

Still, I'm at a loss to understand what you're describing. Except for this habit, my vision is perfectly fine.

What I do experience very strongly - though it is unrelated to this topic - is a very strong tendency to see colours as 3-D layers. Red recedes several millimetres, while blue advances several millimetres. It is inadvertent. I cannot not see this. There's a name for this effect.

Our post boxes just fuhREAK me out!

 

[Tom.G]

Now that you mention it, I too noticed the 3D effect in the mailbox image. I suspected it had something to do with brightness, so investigated.
Converting to Grey scale reduced but did not eliminate the effect. Also the Red has a wider dynamic range of brightness, 0 - 100%. Whereas the Green ranges 0- 74% and Blue ranges 0- 80%.

Maybe we're just built that way with individual sensitivities to the effect.

 

[sophiecentaur]

Still, I'm at a loss to understand what you're describing.

Thanks for that reply Dave; it sort of confirms my thesis. It also shows that binocular vision is far from essential for most human lifestyles. I'm not surprised that you haven't actually been able to see 'the effect' because you have not been using binocular vision, which would be needed for you to identify the planar layout of a TV screen at the level that your brain would need. You have missed out very little with your vision if the only time it's been needed has been to produce an optical illusion!

Vision, along with hearing, sense of colour, taste and touch, is a very 'private' thing. No one can be sure of how another person is perceiving the World. It's only when we use sophisticated tests on many individuals or when we try to reproduce the senses in machines that we can approach anything like an understanding of these things.
This is why I get so cross when people (even on the well-informed and deep-thinking PF) try to dismiss our perception with over simplified models.

PS to @Tom.G too. The red post box effect also make me unwell. I suggest that it could be because of the Chromatic Aberration of our optical system which our processing mostly manages to compensate for. Also, the acuity of red and blue vision is lower than the luminance, which is sensitive to the mid range of frequencies. The post box pattern could be one of those things that Evolution never prepared us for but it wouldn't surprise me to find that the effect is used by brightly coloured birds and insects, to make them harder to catch by (binocularly advantaged) predators. Evolution is responsible for everything.

 

[JT Smith]

My problem is that there is an implication that, viewing with one eye should somehow prevent you from or detract from getting 3D information of a scene...

...The title of the thread implies that this is a big surprise but why should it be?

Because that's the common reaction. People are surprised.

When you're looking a real scene and you close one eye there is usually a sense of things flattening out, a loss of 3D perception. But the opposite happens with a photograph. It surprises people.