![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
||||||
![]() |
|||||||||||
![]() |
|
||||||||||
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
Perceiving Stereoscopic Images |
|||||||||||||||||
![]() |
Until relatively recently, mankind was not aware that there was a separable binocular depth sense. Through the ages, people like Euclid and Leonardo understood that we see different images of the world with each eye. But it was Wheatstone who in 1838 explained to the world, with his stereoscope and drawings, that there is a unique depth sense, stereopsis , produced by retinal disparity. Euclid, Kepler, and others wondered why we don't see a double image of the visual world. Wheatstone explained that the problem was actually the solution, by demonstrating that the mind fuses the two planar retinal images into one with stereopsis ("solid seeing"). 1 A stereoscopic display is an optical system whose final component is the human mind. It functions by presenting the mind with the same kind of left and right views one sees in the visual world. |
||||||||||||||||
![]() |
Retinal Disparity Try this experiment: Hold your finger in front of your face. When you look at your finger, your eyes are converging on your finger. That is, the optical axes of both eyes cross on the finger. There are sets of muscles which move your eyes to accomplish this by placing the images of the finger on each fovea, or central portion, of each retina. If you continue to converge your eyes on your finger, paying attention to the background, you'll notice the background appears to be double. On the other hand, when looking at the background your eyes are converging on it, and your finger, with introspection, will now appear to be double. If we could take the images that are on your left and right retinae and somehow superimpose them as if they were Kodachrome slides, you'd see two almost-overlapping images - left and right perspective viewpoints - which have what physiologists call disparity . Disparity is the distance, in a horizontal direction, between the corresponding left and right image points of the superimposed retinal images. The corresponding points of the retinal images of an object on which the eyes are converged will have zero disparity. |
||||||||||||||||
![]() |
Retinal disparity is caused by the fact that each of our eyes sees the world from a different point of view. The eyes are, on average for adults, two and a half inches or 64 millimeters apart. The disparity is processed by the eye-brain into a single image of the visual world. The mind's ability to combine two different, although similar, images into one image is called fusion, and the resultant sense of depth is called stereopsis . |
||||||||||||||||
![]() |
Parallax A stereoscopic display is one that differs from a planar display in only one respect: It is able to display parallax values of the image points. Parallax , as we shall see, produces disparity in the eyes, thus providing the stereoscopic cue. The stereoscope is a means for presenting disparity information to the eyes. In the century and a half since Wheatstone, it has been improved and, as mentioned in the preface, the View-Master is one such device. In the View-Master stereoscope there are two separate optical systems. Each eye looks through a magnifying lens to see a slide. The slides are taken with a pair of cameras mounted on a base, to replicate the way we see the world with two eyes. |
||||||||||||||||
Each camera sees a different perspective view of the world because it is offset horizontally from the other. When the two slides are viewed in the View-Master the result is the perception of an image with stereoscopic depth, resulting from the parallax information incorporated in the two photos. The parallax values of image points in the pair of slides may be measured just as we conceived of measuring retinal disparity - by laying the slides on top of each other and using a ruler between corresponding points. |
|||||||||||||||||
![]() |
|||||||||||||||||
Electro-stereoscopic displays provide parallax information to the eye by using a method related to that employed in the stereoscope. In the CrystalEyes or SimulEyes display, the left and right images are alternated rapidly on the monitor screen. When the viewer looks at the screen through shuttering eyewear, each shutter is synchronized to occlude the unwanted image and transmit the wanted image. Thus each eye sees only its appropriate perspective view. The left eye sees only the left view, and the right eye only the right view. |
|||||||||||||||||
![]() |
If the images (the term "fields" is often used for video and computer graphics) are refreshed (changed or written) fast enough (often at twice the rate of the planar display), the result is a flickerless stereoscopic image. This kind of a display is called a field-sequential stereoscopic display. It was this author's pleasure to build, patent, and commercially exploit the first flickerless field-sequential electro-stereoscopic display, which has become the basis for virtually every stereoscopic computer graphics product. 2 |
||||||||||||||||
When you observe an electro-stereoscopic monitor image without our eyewear, it looks like there are two images overlayed and superimposed. The refresh rate is so high that you can't see any flicker, and it looks like the images are double-exposed. The distance between left and right corresponding image points (sometimes also called "homologous" or "conjugate" points) is parallax, and may be measured in inches or millimeters. Parallax and disparity are similar entities. Parallax is measured at the display screen, and disparity is measured at the retinae. When wearing our eyewear, parallax becomes retinal disparity. It is parallax which produces retinal disparity, and disparity in turn produces stereopsis. Parallax may also be given in terms of angular measure, which relates it to disparity by taking into account the viewer's distance from the display screen. Because parallax is the entity which produces the stereoscopic depth sensation, it's important to study it. Below is a classification of the kinds of parallax one may encounter when viewing a display screen Parallax Classifications |
|||||||||||||||||
![]() |
Four basic types of parallax are shown in the drawings. In the first case, zero parallax, the homologous image points of the two images exactly correspond or lie on top of each other. When the eyes of the observer, spaced apart at distance t (the interpupillary or interocular distance, on average two and a half inches [63 mm]), are looking at the display screen and observing images with zero parallax, the eyes are converged at the plane of the screen. In other words, the optical axes of the eyes cross at the plane of the screen. When image points have zero parallax, they are said to have zero parallax setting (ZPS). |
||||||||||||||||
![]() |
The drawing illustrates uncrossed or positive parallax . In one type of positive parallax, the axes of the left and right eyes are parallel. This happens in the visual world when looking at objects a great distance from the observer. For a stereoscopic display, when the distance between the eyes t equals the parallax t , the axes of the eyes will be parallel, just as they are when looking at a distant object in the visual world. We'll learn that having parallax values equal to t , or nearly t , for a small screen display will produce discomfort. |
||||||||||||||||
Any uncrossed or positive value of parallax between t and zero will produce images appearing to be within the space of the cathode ray tube (CRT), or behind the screen. We will say that such objects are within CRT space . |
|||||||||||||||||
![]() |
Another kind of positive parallax is shown here, in which images are separated by some distance greater than t . In this case, the axes of the eyes are diverging . This divergence does not occur when looking at objects in the visual world, and the unusual muscular effort needed to fuse such images may cause discomfort. There is no valid reason for divergence in computer-generated stereoscopic images. |
||||||||||||||||
![]() |
In the drawing, we give the case in which the eyes' axes are crossed and parallax points are said to be crossed, or negative. Objects with negative parallax appear to be closer than the plane of the screen, or between the observer and the screen. We say that objects with negative parallax are within viewer space. |
||||||||||||||||
Interaxial Separation |
|||||||||||||||||
![]() |
The distance between the lenses used to take a stereoscopic photograph is called the interaxial separation , with reference to the lenses' axes. The axis of a lens is the line which may be drawn through the optical center of the lens, and is perpendicular to the plane of the imaging surface. Whether we're talking about an actual lens or we're using the construct of a lens to create a computer-generated image, the concept is the same. If the lenses are close together, the stereoscopic depth effect is reduced. In the limiting case, the two lenses (and axes) correspond to each other and a planar image results. As the lenses are located farther apart, parallax increases, and so does the strength of the stereoscopic cue. |
||||||||||||||||
Congruent Nature of the Fields The left and right image fields must be identical in every way except for the values of horizontal parallax. The color, geometry, and brightness of the left and right image fields need to be the same or to within a very tight tolerance, or the result will be "eyestrain" for the viewer. If a system is producing image fields that are not congruent in these respects, there are problems with the hardware or software. You will never be able to produce good-quality stereoscopic images under such conditions. Left and right image fields congruent in all aspects except horizontal parallax are required to avoid discomfort.. 3 Accommodation/Convergence Relationship The eyes' axes will be converged as if portions of a stereoscopic image are at different distances, but they remain focused ( accommodated ) on the plane of the screen. This is the only significant manner in which an electro-stereoscopic display differs from the way we see objects in the visual world. In the example given earlier (page 7), in which your eyes were asked to converge on your finger, they will also focus on the finger. When the eyes converge on the background, they will also focus on the background. This accommodation/convergence relationship is a habitual response learned by everyone: Invariably, in looking at objects, accommodation and convergence correspond. However, looking at an electro-stereoscopic image results in an exception to this relationship. Because the action of the muscles controlling convergence and the muscles controlling focusing depart from their habitual relationship, some people may experience an unpleasant sensation when looking at stereoscopic images, especially images with large values of parallax. Thus, attention needs to be paid to parallax. Experience teaches that it's better to use the lowest values of parallax compatible with a good depth effect in order to minimize the breakdown of accommodation/ convergence. Low parallax values will help to reduce viewer discomfort. |
|||||||||||||||||
To clarify the concept, look at the sketch which gives the case of ZPS (zero parallax setting) for image points, in which the eyes are converged and accommodated on the plane of the display screen. This is the only case in which the breakdown of accommodation/ convergence does not occur when looking at a projected plano-stereoscopic display. (An electronic display screen may be considered to produce a rear-projected image, in contrast to a stereoscope display which places the two views in separate optical systems.) Accordingly, low values of parallax will reduce the breakdown of accommodation/convergence, and large values of parallax will exacerbate it. This subject will be discussed in greater detail in the next chapter. |
|||||||||||||||||
![]() |
|||||||||||||||||
Control of Parallax |
|||||||||||||||||
![]() |
The goal when creating stereoscopic images is to provide the deepest effect with the lowest values of parallax. This is accomplished in part by reducing the interaxial separation. In only one case, the ZPS condition, will there be no breakdown of the accommodation/ convergence relationship. If the composition allows, it is best to place the principal object(s) at or near the plane of the screen, or to split the difference in parallax by fixing the ZPS on the middle of an object so that half of the parallax values will be positive and half negative. As a rule, don't exceed parallax values of more than 1.5 degrees. 4 That's half an inch, or 12 millimeters, of positive or negative parallax for images to be viewed from the usual workstation distance of a foot and a half. This conservative rule was made to be broken, so let your eyes be your guide. Some images require less parallax to be comfortable to view, and you may find it's possible to greatly exceed this recommendation for negative parallax values or viewer-space effects. Expressing parallax in terms of angular measure directly relates it to retinal disparity. For example, a third of an inch of parallax produces the same value of retinal disparity at three feet as two-thirds of an inch of parallax at six feet. Keep in mind the distance of viewers from the screen when composing a stereoscopic image. |
||||||||||||||||
Crosstalk Crosstalk in a stereoscopic display results in each eye seeing an image of the unwanted perspective view. In a perfect stereoscopic system, each eye sees only its assigned image. That's what happens in a stereoscope, because each slide is optically separated from the other, as shown on page 8. Our technology, although good, is imperfect. In particular, there are two opportunities for crosstalk in an electronic stereoscopic display: departures from the ideal shutter in the eyewear, and CRT phosphor afterglow.. 5 If the electro-optical shutters were ideal, they would block all of the unwanted image. The shutters in our products are so good that no unwanted image will be perceived because of incomplete shutter occlusion. All the crosstalk is produced by another phenomenon - the afterglow of the display monitor's CRT phosphors. In an ideal field-sequential stereoscopic display, the image of each field, made up of glowing phosphors, would vanish before the next field was written, but that's not what happens. After the right image is written, it will persist while the left image is being written. Thus, an unwanted fading right image will persist into the left image (and vice versa). The term ghosting is used to describe perceived crosstalk. Stereoscopists have also used the term "leakage" to describe this phenomenon. The perception of ghosting varies with the brightness of the image, color, and - most importantly - parallax and image contrast. Images with large values of parallax will have more ghosting than images with low parallax. High-contrast images, like black lines on a white background, will show the most ghosting. Given the present state of the art of monitors and their display tubes, the green phosphor has the longest afterglow and produces the most ghosting. If an image is creating ghosting problems, try reducing the green in the image. Curing the Problems The breakdown of the accommodation/convergence relationship and ghost images produced by crosstalk can detract from the enjoyment of a stereoscopic image. Fortunately, the cure for both problems is the same: Use the lowest values of parallax to produce a good stereoscopic effect. As corresponding points approach the ZPS, both problems go away! As corresponding points move closer together, accommodation and convergence approach their usual relationship and the ghost image also diminishes, since it will increasingly correspond with the wanted image. Summary We have seen that a stereoscopic display differs from a planar display by incorporating parallax . Parallax is classified into positive parallax , which produces an image distant from the viewer (in CRT space), and negative parallax , which produces off-screen effects or images in viewer space. The perceptual artifacts introduced by the limitations of display tubes and the human visual system may be held to tolerable limits by properly adjusting the parallax, a subject which will be addressed further in the next chapter. References 1. Wheatstone, Charles. On some remarkable, and hitherto unobserved, phenomena of binocular vision (Part the first). Philosophical Transactions of the Royal Society of London , 1838, 371-94. 2. Lipton, Lenny, et al. Stereoscopic Television System . U.S. Patent No. 4,523,226, June 11, 1985. 3. Lipton, Lenny. Binocular symmetries as criteria for the successful transmission of images. Processing and Display of Three-Dimensional Data II , SPIE Vol.507, 1984. 4. Valyus, N.A. Stereoscopy . London, New York: Focal Press, 1962. 5. Lipton, Lenny. Factors affecting ghosting in a time-multiplexed plano-stereoscopic CRT display system. SPIE Proceedings , Vol.761, 1987. |
|||||||||||||||||
![]() |
![]() |
All materials © Copyright 1996-97, StereoGraphics Corporation. |
|