Friday, August 5, 2011

Principles of fusion

When I first heard about binocular fusion, I thought it was a way to combine two pictures in order to make a third picture. After all, I thought, we can only see a picture. Left eye sees a picture, right eye sees a picture, the result is something we see, thus the result is also a picture.

If you are not quite sure how this line of thought is possible, think how people thought about motion before Galileo. Galileo probably had binocular vision, since he invented this boat argument about relative motion to explain how it could seem to us that the Earth is not moving, while it is - I thought about this quite a bit after reading Paul Feyerabend's "Against method". On a related note, now this book also allows me to understand how, for example, the Bates method or the Eyebody method could truly correct the refractive error in the eye and thus truly treating myopia, presbyopia, astigmatism, even though there were multiple experiments that convincingly proved that those things cannot be treated.

As I learned from Feyerabend, it had been known long Galileo that the Earth was not moving. The Earth was still in one place, and there was a proof. The proof was via the so-called "Tower argument". Go to the top of the tower, drop some object. The object falls near the base of the tower. Therefore, the Earth is not moving.

This argument is not wrong. It just exists inside a certain theory. Outside of this theory the argument is not wrong, it is almost meaningless. For a person with normal binocular vision it is just as meaningless to think of binocular vision as "the picture". When we have a conversation of a person without binocular experience and  a person with binocular vision, they indeed speak a different language, and what happens is, in the words of W.V.O.Quine, a radical translation. There is no point asking if a person without binocular vision has depth perception - the words "depth perception" have a totally different meaning.

It was only today that I understood just how wrong my understanding of fusion was. I was practicing fixing my gaze and holding it still, largely because of what I read in Susan Barry's book "Fixing my faze". Yesterday I thought I did fine. However, today I noticed that as I was trying to fix my gaze, sometimes I saw a double image. Sometimes I saw alternating switching. Furthermore, in case of one particular word that I read on some label, the first several letters were a little above, the last several letters were a little below, and the middle letter was alternating up and down, moving in line either with the first or with the last letters. It felt to me that I was almost fusing these letters and I only had to wait until the mind gets tired of alternating and decides finally on some picture in between - just like after the mind fuses a finger on the left and a finger on the right, it is suppose to see a finger in between.

It is as if there were a textbook for fusion. In the beginning there would be images for the left and the right eye to fuse. In the end there would be answers. Not like in the "Magic Eye" books, but the real answers. The end of the book would show the fused images. This is how I was thinking about it.

Binocular fusion is always and inherently three-dimensional.

Fusion is something that happens mentally, in the brain. Take an apple, for example. You can smell the apple, and the smell comes from particles acting on your receptors. Replace your nose by somebody else's nose, and it would smell the same particles. Similarly, take the test. Replace your tongue by somebody else's tongue, and there would be essentially the same sensory input. However, the fusion of small+taste happens mentally, in the brain. Two different people will perform this fusion differently. It does not exist in any objective sense, as opposed to the sound waves, the light waves, the particles that we can smell or taste.

Binocular vision combines two sensory inputs into a sensation that has no physical manifestation. Binocular fusion exists only in the brain. One cannot build a computer that will have "the same" binocular vision of an apple as a person, but one can build an array of photoreceptors similar to a retina that will, in a sense, have "the same" monocular vision as that of a person.

Therefore, it is not very useful to keep switching between the left-eye and right-eye images, hoping that they will somehow fuse. Fusion is not a combination of images that you can do using some computer program such as Adobe Photoshop. Even the idea of "seeing" fusion is confusing and ambiguous.

Fusion is not "seen". It is sensed or experienced. We just call it seeing. It may be more precise to call it 2-seeing, or f-seeing, or "seeing with binocular vision". This is a different sensory act. Last weekend I had a clear distinction between 2d-memory and 3d-memory, between imagining things in 2d (on the left eye or on the right eye or across both) or imagining things in 3d. (That is, binocular vision is a different sense, even though it is built on top of another sense, monocular vision.) However, I am not quite sure if I can make such distinctions between 2d and 3d any more.

Stereopsis comes before all other aspects of binocular vision. Look around. If there is some object that you are only seeing with one eye because of, say, a convergence deficiency, this object will look flat, precisely because you are seeing it with one eye. Convergely, if you are seeing an object that is flat, it means there is no fusion, because the fusion is always three-dimensional. If you are seeing an object with only one eye, the other eye may get misaligned. Similarly, even if the eyes are aligned, this does not mean that binocular fusion will happen. However, if binocular fusion does happen and you are seeing the object in 3d, the eyes will be aligned. If the object is moving, as long as you can keep the image three-dimensional, you can follow it with one eye, and both eyes will follow. It is difficult to control both eyes to follow an object, it is easier to keep an intent for the object to look three-dimensional, and then follow it somehow, without specifically trying to use both eyes.
If you see double vision or alternation, there is no fusion. If you are reading a book and you do not see with both eyes that the book is flat, then there is no fusion, and you are only using one eye. Don't worry, the image is probably there, on both retinas. Assuming for now that it is on properly corresponding parts of the retinas, you "just" need to see one instead of two, but out of being. You can create the right circumstances, but then you can only let go and let binocular fusion happen - or not happen. Ideally, without attachment to the outcome.

No comments:

Post a Comment