Japanese Psychological Research 2012, Volume 54, No. 1, 54–70 Special issue: Stereoscopic depth perception
doi: 10.1111/j.1468-5884.2011.00505.x Review
Early studies of binocular and stereoscopic vision1 NICHOLAS J. WADE2* HIROSHI ONO
University of Dundee
York University
Abstract: The revolution in binocular vision (in the 1830s) was occasioned by Wheatstone’s invention and application of the stereoscope to demonstrate depth from retinal disparity. The stereoscope, perhaps more than any other instrument, ushered in the era of experimentation to vision. It fulfilled the scientific desire to examine binocular vision by observation and experiment. The stereoscope is a simple optical device that presents slightly different figures to each eye. If these figures have appropriate horizontal displacements or disparities then depth is seen. Wheatstone achieved for space perception what Newton had for color vision: the phenomena could be removed from their object base. Newton’s decomposition of white light into its spectral components removed the perception of color from the colored objects that naturally conveyed it. Wheatstone’s decomposition of stereoscopic depth into its disparate projections to each eye removed the perception of depth from the solid objects that naturally conveyed it. Color and depth could be examined in the laboratory, and the methods of the natural sciences could be applied to their investigation. Key words: stereoscope, binocular vision, depth perception, Wheatstone.
The introduction of the stereoscope inaugurated a new epoch in the physiology of vision, opened a wide field for further inquiry, and suggested additional methods of investigation, while the theory of binocular vision has been greatly modified by results which have been obtained through the medium of the instrument. (Towne, 1862, p. 70) The study of vision has been advanced by the invention of instruments that control the stimuli presented to an observer. This enabled the methods of physics, where a single stimulus variable is manipulated and others are kept constant, to be applied to visual phenomena. In the
context of color vision, Newton (1704) was able to select parts of the spectrum (by passing white light through a prism) and combine them with other parts in systematic ways. By this means, some of the rules of color combination could be established as well as the nature of light itself. For the first time, color vision could be investigated without using colored objects. It could be argued that Newton separated color perception from its object base, and changed the ways in which it was studied and interpreted. The laboratory, rather than the natural environment, provided the scene for studying the seen. The same argument can be made for the stereoscope in the context of space perception. Following its
*Correspondence concerning this article should be sent to: Nicholas J. Wade, School of Psychology, University of Dundee, Nethergate, Dundee DD1 4HN, UK. (E-mail:
[email protected]) 1
This research was supported in part by Grant A0296 from the Natural Sciences and Engineering Research Council of Canada to H. Ono.
2
The authors wish to thank Linda Lillakas for her helpful comments and discussion on earlier versions of this paper.
© 2012 Japanese Psychological Association. Published by Blackwell Publishing Ltd.
Binocular and stereoscopic vision
invention, by Wheatstone in the 1830s, depth could be related to retinal disparity and it could be examined in the laboratory using the methods of physics. Paired drawings or photographs could be presented separately to each eye and the ensuing depth perceived could be determined. The stereoscope freed space perception from its object base, and it could be examined with the precision that had been applied to color vision. However, the road to the discovery of the link between disparity and stereoscopic depth was long and less than linear. We present an account of this journey, from ancient times to Wheatstone’s instrumental invention and beyond. The journey is not complete, but appreciating the turns it has taken in the past can only be of benefit to present day travelers along this visual path. In a historical sense, the distinction between monocular and binocular vision derived from the occasional experience of diplopia when using two eyes, and from the mistakes made by those with only one eye. For example, Aristotle (384–322 B.C.) described how one object could be seen double when one eye was moved by the finger (Ross, 1931): “if a finger be inserted beneath the eyeball without being observed, one object will not only present two visual images, but will create an opinion of its being two objects” (pp. 461b–462a). The reports of errors in reaching after one eye had been lost came much later. Boyle (1688) made such an observation: “Haveing frequently occasion to pour Distill’d Waters and other Liquors out of one Vial into another, after this Accident [of losing an eye] he often Spilt his Liquors, by pouring quite Beside the necks of the Vials he thought he was pouring them directly Into” (p. 255). Moreover, the perceptual problems associated with strabismus were also rather later in being examined (Wade, 1998). Thus, interest was directed either to departures from single vision or to errors made in depth perception. The characteristic of perception that was considered normal, and corresponded to common experience, was single vision with two eyes. However, most commentators did not examine how this came about and considered that monocular vision was superior to binocu-
55
lar; this was because vision with one eye was not prone to the vagaries of diplopia. The long-held superiority of monocular vision was supported by the prevailing anatomy and speculative physiology derived from Galen (ca. 130–200) in the 2nd century. He addressed matters of vision in the context of anatomy, though he made many astute observations, particularly in the context of binocular vision. His theory of vision was physiological: the visual spirit passed along the hollow tubes of the optic nerves to interact with returning images of external objects in the crystalline lens. By adopting an anatomical and physiological analysis of vision, Galen was confronted with the existence of two eyes and the observation by them of a single visual world.The visual spirit originated in the ventricles and was united at the chiasm only to be separated again to pass to each eye. According to this view, when one eye alone was used all the visual spirit could be passed to it, rather than being shared between the eyes. This conception was rarely challenged in antiquity, despite the fact that both Ptolemy (ca. 100–175) and Ibn al-Haytham (or Alhazen, ca. 965–1039) proposed that vision with two eyes was better than with one. As with many other aspects of visual perception, binocularity has been analyzed in terms of physical optics. More than 2000 years ago, Euclid (ca. 323–283 B.C.) examined binocular vision with the consistency that he had adopted for other aspects of spatial vision and found that it could be reduced to optical projections. In fact, his discussion of the projections from two eyes was rather cursory, being restricted to three different sizes of sphere with respect to the interocular distance. Diagrams derived from his writings are shown in Figure 1. Euclid’s analysis was geometrical: he examined the three dimensions of a sphere that could be observed by two eyes, and simply related them to the amount of the spheres that would be seen. Euclid’s use of a sphere was to have unexpected implications, because when Leonardo da Vinci (1452–1519) examined binocular projections to the eyes he also used a sphere as the stimulus (Figure 1). Leonardo was the principal proponent of binocular superiority. He © Japanese Psychological Association 2012.
56
N. J. Wade and H. Ono
compared binocular vision of a scene with a painting of it, and noted that the painting did not result in seeing the relief or depth of the scene. He made many drawings of projections to one or two eyes of a small sphere, indicating how all the background could be seen with two eyes but not with one (see Wade, Ono, & Lillakas, 2001). The use of a sphere with a diameter less than the separation between the eyes reflected Leonardo’s return to the situation examined by Euclid. Indeed, as Wheatstone (1838) ruefully noted, had Leonardo chosen any object other than a sphere he might have realized and represented binocular disparities. Another aspect of using two eyes that exercised students after Leonardo was possible differences between them. This became the issue of eye dominance, and Porta (1593) described a test for it. It is a sighting test in which a distant object viewed with both eyes is aligned with a hand-held stick (Figure 2). Porta
reported that the right eye is used for this task, and he also stated that the right eye is preferred when different patterns are viewed by two eyes. The notion of rivalry between the two eyes was also given empirical support by Porta. When considering singleness of vision with two eyes, he adopted the theory that we only use one at a time, and this could be simply demonstrated: Nature has given us two eyes, one on the right and the other on the left, so that if we are to see something on the right we use the right eye, and on the left the left eye. It follows that we always see with one eye, even if we think both are open and that we see with both. We may prove it by these arguments: To separate the two eyes, let us place a book before the right eye and read it; then someone shows another book to the left eye, it is impossible to read it or even see the pages, unless for a
Figure 1 The upper three figures are representations made from Euclid’s Optics showing two eyes viewing a sphere that has a diameter equal to the interocular separation or larger and smaller than it (Euclid, 1895). Leonardo da Vinci similarly used a sphere, with a diameter less than the interocular separation, to examine binocular vision and to compare it with monocular observation. The remaining drawings are from his Notebooks (Wade et al., 2001). © Japanese Psychological Association 2012.
Binocular and stereoscopic vision
short moment of time the power of seeing is taken from the right eye and borrowed by the left. (Porta, 1593, pp. 142–143) Porta’s view became known as suppression theory. The contrary view, that we fuse or combine the images from each eye, was proposed soon after by Aguilonius (1613). He coined the term horopter to describe the plane in which corresponding points in the two eyes were stimulated. Since the early 17th century, the study of binocular vision has often been
57
seen as a contrast, even a conflict, between suppression and fusion theories.
Methods of studying binocular vision before the stereoscope The essence of investigating binocular vision was distilled from the methods adopted for stimulating the two eyes. Rubens’s engraving (Figure 2, lower) demonstrated the technique of fixating on one object located further from
Figure 2 Two illustrations from Aguilonius (1613), both of which were designed by Peter Paul Rubens. The upper engraving (frontispiece from Book III) is possibly depicting the cosmic observer performing Porta’s sighting test; having pointed to the stick, held by a putto, the observer closes the left eye to determine whether the finger is still aligned with the stick. The lower engraving (frontispiece to Book IV) shows the cosmic observer fixating the central cross (on the screen), thus producing crossed visible directions of the near object. The putti are pointing to the discs on the screen which mark the locations of the crossed directions. © Japanese Psychological Association 2012.
58
N. J. Wade and H. Ono
the eyes than another. This method was introduced by Ptolemy (Smith, 1996), and elaborated by Alhazen (1572; Sabra, 1989), before its widespread adoption in the 17th and 18th centuries. Another technique involved placing a septum between the eyes, so that peripheral objects could be seen by one eye but not the other. Galen described this method, and it was pursued by Porta and Aguilonius. Observing distant objects through a small aperture, so positioned that they are aligned each with one eye, was used by Le Clerc (1712) and Desaguliers (1716) to examine binocular combination, whereas Du Tour (1760) placed shapes or patterns on different sides of a septum (Figure 3). However, the over-riding interest was not in binocular depth perception but in examining binocular single vision and binocular rivalry.
Moreover, the general source of interest was not space perception but color vision: would different colors presented to corresponding regions of the eyes combine or compete? In this context, Du Tour (1760) provided a clear description of binocular competition. He placed a board between his eyes and attached blue and yellow fabric in equivalent positions on each side, or the fabric was placed in front of the fixation point. When he converged his eyes to look at them they did not mix but alternated in color, confirming the observation of Desaguliers (1716). Du Tour (1761) also examined binocular contour rivalry: he held a prism in front of one eye, and produced the clearest early description of binocular competition. Either the stimulus presented to one or the other eye would be visible, or some mixture of the two views would present itself:
Figure 3 18th century techniques for presenting different stimuli to the two eyes. Lower left, Le Clerc’s (1712) method of viewing through an aperture nearer than the targets. A similar technique was used by Desaguliers (1716) as shown in the upper figure. Du Tour (1760) placed targets on either side of a septum (lower right); he also used prisms to stimulate the eyes with different patterns. © Japanese Psychological Association 2012.
Binocular and stereoscopic vision
If one applies a prism held vertically before one of the eyes, so only refracted rays of light are passed to that eye, and with the other eye open, it is certain that different objects will be projected on corresponding portions of the two retinas. . . sometimes I would see only objects projected in the bare eye, sometimes only those in the eye covered by the prism, and sometimes the objects projected in one would seem to me to intermingle with the objects projected in the other. (p. 500) Singleness was at the heart of Descartes’s (1637/1965, 1664/1972) analysis of binocular vision: he posited its singleness in a singular organ in the brain, the pineal body. Corresponding points on each retina projected to the same locations on the pineal body, thus defining single vision. The spur to studying binocular vision by his opponents was to attack this theory. For example, Le Clerc (1679) argued that retinal disparity would not yield single vision: Monsieur Descartes having considered that according to his principles external objects
59
should make an impression on both eyes, and that the soul nevertheless had only one perception believed that the images of the same object found in the two eyes are reunited in the brain; but if this great genius had reflected a little more on the demonstrations which he gave in his Treatise on Man, he would have recognized that the images in the two eyes although produced by the same object, are different, and because of these differences their reunion is impossible. (pp. 44–46) Le Clerc, who was an authority on perspective, provided clear diagrams of retinal disparities (Figure 4) but he did not see the link between disparities and depth; rather he used disparities to disparage Descartes’s theory. With regard to Figure 4 left, Le Clerc (1679) wrote: It is evident that the images in the two eyes are different from one another; for supposing that the object DEFG is seen by the two eyes A and B, the eye A sees G between the rays AF and AE, the eye B on the other hand sees
Figure 4 Two of Le Clerc’s (1679) diagrams of retinal disparities. They were used to argue against Descartes’s theory of single vision rather than to suggest that disparities could be utilized in vision. © Japanese Psychological Association 2012.
60
N. J. Wade and H. Ono
it between BD, BF and therefore if the eye A sees the point G in the line EF at point H, the eye B sees it in the line FD at point I. (p. 46) Knowledge about retinal disparities was not a novelty for Le Clerc but he represented it in a compelling manner. Differences between images at each eye were clearly described by Ptolemy and Galen. Euclid gave descriptions of the optical projections to a sphere, and this example was pursued by Leonardo for a sphere with a diameter smaller than the interocular separation. However, Leonardo did not restrict his consideration to the sphere alone, but also to the amount of background that would be visible: with one eye part of the background is always obscured from view, but this does not necessarily happen with binocular viewing. Wheatstone reflected on Leonardo’s choice of a sphere, suggesting that his expertise in perspective would have led him to realize projective disparity if a cube had been used; Leonardo might also have appreciated the significance of disparate images. Although Wheatstone declared himself unaware of any subsequent attempt using a cube, he could not have known about Le Clerk’s studies. Le Clerc was both an artist and an academician who wrote on perspective. In his two books on vision (published in 1679 and 1712) he described and illustrated the binocular projections from a cube, as well as other objects, and he discussed the differences in the angles subtended by a given side to each eye, as is evident from Figure 4. Wheatstone’s genius was not to describe disparity but to demonstrate the uses to which it could be put: depth or distance perception.
Visual distance before the stereoscope Distance was known to be involved not only in the perception of objects in space, but also in pictorial representations of them. Moreover, distance was described in terms of a variety of cues that could assist in determining whether one object was nearer or farther than another. These cues to distance were repeated, with varying emphases, up to the early 19th century, © Japanese Psychological Association 2012.
when Wheatstone (1838) added that of retinal disparity.Wheatstone placed binocular vision in the domain of space perception as well as singleness of vision, and in so doing introduced distance as a relevant dimension. Space perception represents an arena in which optics and observation were often in conflict. In his Optics, Euclid (1895) provided geometrical analyses of many spatial phenomena. Thus, for example, size perception was equated with visual angles, so that the same object would have a different apparent size according to its distance from the eye. Ptolemy’s interpretations were more subtle, because he realized that visual angles alone did not accord with the characteristics of observation. By adding distance and orientation to visual angles, he was able to give accounts of size and shape constancies. Alhazen pursued this Ptolemaic line, which was given further elaboration by Descartes’s (1637/1965): As to the manner in which we see the size and shape of objects. . . their size is estimated according to the knowledge, or the opinion, we have of their distance, compared with the size of the images that they imprint on the back of the eye; and not absolutely by the size of these images, as is obvious enough from this: while the images may be, for example, one hundred times larger when the objects are quite close to us than when they are ten times farther away, they do not make us see the objects as one hundred times larger because of this, at least if their distance does not deceive us. And it is also obvious that shape is judged by the knowledge, or opinion, that we have of the position of various parts of the objects, and not by the resemblance of the pictures in the eye; for these pictures usually contain only ovals and diamond shapes, yet they cause us to see circles and squares. (p. 107) If the retinal size can be rescaled according to the knowledge we have of the distance of objects then some sources of this knowledge must be available. These became known as cues to distance. Alhazen speculated that familiarity with objects assisted in the perception of their distance, and thereafter increasing concern was
Binocular and stereoscopic vision
given to a multiplicity of cues to distance. Descartes specified two others beside familiar size, convergence and accommodation, and he defined the distances over which they operated. Rohault (1671) added interposition and motion parallax to these. de La Hire (1694) stated that motion parallax was the most important cue, a view with which Wheatstone (1838) concurred (Ono & Wade, 2005). In stating that perceived distance is inferred, Berkeley (1709) argued that (for relatively near objects) the inference was based on convergence, accommodation, and image clarity. Most particularly, the motion of the eyes, in convergence, was of principal importance, and his recourse to motor mechanisms was to influence many 18th century analyses of space perception. At the beginning of the 19th century, Young (1807) summarized the general understanding about the perception of distance: We estimate distances much less accurately with one eye than with both, since we are deprived of the assistance usually afforded by the relative situation of the optical axes. . . Our idea of distance is usually regulated by a knowledge of the real magnitude of an object, while we observe its angular magnitude; and on the other hand a knowledge of the real or imaginary distance of the object often directs our judgment of its actual magnitude. (p. 453) However, it was the idea of inference that was to prove potent in the later decades of the century. Euclid’s geometrical analysis of size perception did have the virtue of accuracy, and so it is not surprising that it reappeared centuries later in the context of artistic representation. The Renaissance mathematician, Leon Battista Alberti (1435/1966) wrote a treatise, On Painting, which described the then new technique of linear perspective. In addition to his mathematical treatment of perspective, he devised a simple method for capturing visual angles: with the eye in a fixed position, the objects in a scene can be traced on a window pane. This has become called Alberti’s window, although Leonardo’s name is often given to it following
61
his description of the same procedure. In many ways, perspective was a formalization of Euclid’s optics, as it is concerned with capturing visual angles of objects at different distances. Euclid provided an excellent theory of picture production, but not of space perception. His influence was felt by Leonardo, who grappled with the issues of apparent depth (or its absence) in paintings. Nonetheless, it did result in both artists and scientists describing the cues to depth that were present in perspective pictures, such as interposition, relative size, and texture gradients. They are still called painter’s cues to depth.
The stereoscope The stereoscope was invented in the early 1830s, and it opened a new world for the study of binocular vision. That world was the laboratory, and with the aid of the stereoscope the methods of physics could be applied to the investigation of spatial vision. Wheatstone made mirror and prism stereoscopes as early as 1832, but he only described the mirror version in his classic memoir of 1838 (Figure 5). His first stereoscopes were made by the London optical instrument firm of Murray and Heath. The first published account of the stereoscope was in Mayo (1833): One of the most remarkable results of Mr. Wheatstone’s investigations respecting binocular vision is the following. A solid object being placed so as to be regarded by both eyes, projects a different perspective figure on each retina; now if these two perspectives be actually copied on paper, and presented one to each eye, so as to fall on corresponding parts, the original solid figure will be apparently reproduced in such a manner that no effort of the imagination can make it appear as a representation on a plane surface. (p. 288) Wheatstone described the mirror stereoscope at a meeting of the Royal Society of London in June, 1838, and he demonstrated the device to a meeting of the British Association for the © Japanese Psychological Association 2012.
62
N. J. Wade and H. Ono
Advancement of Science held at Newcastle in August, 1838. Stereoscopes were made for him by the London optical firm of Murray and Heath in 1832, and these involved combinations both by reflection and refraction. That is, Wheatstone had prism as well as mirror stereoscopes made for him, but he only described the reflecting instrument in his memoir of 1838 (see Wade, 1983). Brewster devised his lenticular stereoscope in 1849 (Figure 6). It consisted of a single lens cut in half so that the two half-lenses, when appropriately mounted, acted as magnifiers as well as prisms, fusing adjacent stereo drawings or photographs. The first model was made by George Lowdon, an optical instrument maker in Dundee, but the version displayed at the Great Exhibition, held in Crystal Palace, London in 1851 was made by Louis Jules Duboscq of Paris. It was more popular than the mirror stereoscope because it was more compact and could be used more conveniently
with paired photographs. Duboscq took out a patent for his instrument in 1852; it was a slightly modified form of Lowdon’s model, with a ground glass endplate so that both printed and transparent stereophotographs could be observed. Brewster (1844a) interpreted stereoscopic phenomena in terms of visual direction: depth was seen at the location of intersection of lines of direction from each eye (see Ono & Wade, 2012). Brewster (1830) initially presented his theory in a long encyclopedia essay on optics; his theory was virtually unchanged by Wheatstone’s instrument and by the observations he obtained with it. Nonetheless, Brewster’s (1851) optical ingenuity led to a wide variety of methods for combining stereo-pairs, which are illustrated in Figure 6; Dove (1851) also indicated many ways in which dissimilar images could be combined using prisms. All Wheatstone’s papers on binocular vision have been reprinted in Wade (1983), as have those of Brewster.
Figure 5 Illustrations from Wheatstone’s (1838) article describing the stereoscope and stereoscopic vision; they were presented as two Plates (pages of figures) in the article. © Japanese Psychological Association 2012.
Binocular and stereoscopic vision
Wheatstone (1838) was able to manipulate the pictures presented to each eye and observe the depth that was produced. In so doing, he found that: . . . the projection of two obviously dissimilar pictures on the two retinæ when a single object is viewed, while the optic axes converge, must therefore be regarded as a new fact in the theory of vision. It being thus established that the mind perceives an object of three dimensions by means of the two dissimilar pictures projected by it on the two retinæ, the following question occurs: What would be the visual effect of simultaneously presenting to each eye, instead of the object itself, its projection on a plane surface as it appears to that eye? (pp. 372–373) Binocular instruments were in existence long before the stereoscope was invented, as was knowledge of retinal disparities (Blundell, 2011; Crone, 1992; Howard, 2002; Wade, 1987, 1998, 2004). Indeed, Wheatstone (1838) described and illustrated the ways in which
63
different stimuli could be viewed without a stereoscope: by under- and over-convergence, by using two viewing tubes, or by a combination of over-convergence and a septum between the eyes, as shown in Plate 1 of Figure 5. Brewster (1856) called such devices ocular stereoscopes because they did not contain any optical aids like lenses, mirrors, or prisms. Wheatstone was able to dissociate accommodation from convergence, and so did not require the instrument he invented; the stereoscope was devised so that others could view dissimilar pictures with ease. The “new fact in the theory of vision” was the systematic dissimilarities between the two pictures and the depth they induced. Wheatstone realized this as a consequence of a fortuitous observation: When a single candle flame is brought near such a [metal] plate, a line of light appears standing out from it, one half being above, and the other half below the surface; the position and inclination of this line changes with the situation of the light and of the observer, but it always passes through the
Figure 6 Drawings of Brewster’s lenticular stereoscope and its mode of operation (from Brewster, 1856), together with illustrations of his many models for binocular combination (from Brewster, 1851). © Japanese Psychological Association 2012.
64
N. J. Wade and H. Ono
centre of the plate. On closing the left eye the relief disappears, and the luminous line coincides with one of the diameters of the plate; on closing the right eye the line appears equally in the plane of the surface, but coincides with another diameter; on opening both eyes it instantly starts into relief. The case here is exactly analogous to the vision of two inclined lines when each is presented to a different eye in the stereoscope. It is curious, that an effect like this, which must have been seen thousands of times, should never have attracted sufficient attention to have been made the subject of philosophic observation. It was one of the earliest facts which drew my attention to the subject I am now treating. (p. 379) Of Wheatstone’s (1838) 12 paired drawings, 11 were used to demonstrate that stimulation of the two eyes with slightly different pictures could lead to depth perception. The odd one (figure 23 in Figure 5), which involved presenting a thick vertical line to the right eye and a thick inclined one with a thin vertical to the left, was taken to show “that similar pictures falling on corresponding points of the two retinæ may appear double and in different places” (p. 384). He reported that the thick lines combined to be seen in depth and the thin line remained visible and vertical. That is, the vertical lines, falling on corresponding retinal points, were not combined. This observation created such controversy that Hering (1862) referred to it as the Wheatstone Experiment (Ono & Wade, 1985). Wheatstone’s (1852) second article on binocular vision was published 14 years later. He described and illustrated an adjustable mirror stereoscope, a prism stereoscope, and a pseudoscope for reversing disparities. The main purpose of these was to extend the range of conditions under which the two eyes could be stimulated: Under the ordinary conditions of vision, when an object is placed at a certain distance before the eyes, several concurring circumstances remain constant, and they always vary in the same order when the dis© Japanese Psychological Association 2012.
tance of the object is changed. Thus, as we approach the object, or as it is brought nearer to us, the magnitude of the picture on the retinæ increases; the inclination of the optic axes, required to cause the picture to fall on corresponding places on the retinæ, becomes greater; the divergence of the rays of light proceeding from each point of the object, and which determines the adaptation of the eyes to distinct vision of that point, increases; and the dissimilarity of the two pictures projected on the retinæ also becomes greater. It is important to ascertain in what manner our perception of the magnitude and distance of objects depends on these various circumstances, and to inquire which are the most, and which the least influential in the judgements we form. To advance this inquiry beyond the point to which it has hitherto been brought, it is not sufficient to content ourselves with drawing conclusions from observations on the circumstances under which vision naturally occurs, as preceding writers on this subject mostly have done, but it is necessary to have more extended recourse to the methods so successfully employed in experimental philosophy, and to endeavour, wherever it be possible, not only to analyse the elements of vision, but also to recombine them in unusual manners, so that they may be associated under circumstances that never naturally occur. (p. 2) Wheatstone (1852) used the stereoscope with adjustable arms to vary the four circumstances mentioned in the quotation (retinal size, convergence, accommodation, and disparity). He found that: The perceived magnitude of an object, therefore, diminishes as the inclination of the axes becomes greater, while the distance remains the same; and it increases, when the inclination of the axes remains the same, while the distance diminishes. When both of these conditions vary inversely, as they do in ordinary vision when the distance of an object changes, the perceived magnitude remains the same. (p. 3)
Binocular and stereoscopic vision
He applied the pseudoscope to reverse the normal relations between monocular and stereoscopic cues to depth: “With the pseudoscope we have a glance, as it were, into another visible world, in which external objects and our internal perceptions have no longer their habitual relation with each other” (p. 12). He remarked on the difficulty of perceiving reversals of relief with the pseudoscope, and the illuminating conditions that are necessary for such reversal. Both Wheatstone and Brewster were acquaintances of William Henry Fox Talbot, who made public his negative-positive photographic process in the year after Wheatstone’s first article on the stereoscope appeared. In fact the term “photographic” was first used by Wheatstone (Arnold, 1977). He immediately grasped the significance of photographing
65
scenes from two positions, so that they would be seen in depth when mounted in the stereoscope. In 1840, he enlisted Talbot’s assistance to take stereo-photographs for him; when they were sent to him the angular separation of the camera positions used to capture the two views was too large (47.5 deg) and Wheatstone suggested that 25 deg would be more appropriate. Klooswijk (1991) has reprinted a section of Wheatstone’s letter to Talbot, and has himself taken stereo-photographs of the bust Talbot probably employed from camera angles of 47.5, 25.0, and 1.75 deg. However, it was Brewster’s lenticular stereoscope which benefited from the invention of photography, and the stereophotographs that exist of Wheatstone and Brewster were made for the lenticular stereoscope (Figure 7).
Figure 7 Upper, a stereodaguerreotype of the Wheatstone family, taken by Antoine Claudet. Lower, a stereocalotype of Brewster, sitting beside a model of his lenticular stereoscope. © Japanese Psychological Association 2012.
66
N. J. Wade and H. Ono
Binocular vision in the 19th century after the invention of the stereoscope Having established that dissimilar pictures, when viewed in the stereoscope, produce the appearance of depth, Wheatstone conducted a series of systematic manipulations of the figures in order to discover the nature of the relationship. In his first article (Wheatstone, 1838) he demonstrated that the sign of disparity (crossed or uncrossed) determined the relative depth seen (nearer or farther), that there was a limit to the disparity yielding singleness of vision, that eye movements were not involved (because depth was seen in disparate afterimages), and that radically different pictures or colors resulted in rivalry. The article was rapidly translated into German, first in summary (Wheatstone, 1839), and later in a full translation (Wheatstone, 1842); much of the subsequent experimental work on binocular vision was conducted in Germany. James (1890) shrewdly observed that Wheatstone’s paper: . . . contains the germ of almost all the methods applied since to the study of optical perception. It seems a pity that England, leading off so brilliantly the modern epoch of this study, should so quickly have dropped out of the field. Almost all subsequent progress has been made in Germany, Holland, and, longo intervallo, America. (pp. 226–227) Among the theoretical issues raised by stereoscopic depth perception and fusion of binocularly disparate images was the extent to which stereoscopic vision was a consequence of eye movements (Ono, Lillakas, & Wade, 2007). Brücke (1841), Prévost (1843), Brewster (1844a), Towne (1869), and LeConte (1881, 1897) proposed eye movement interpretations, despite Wheatstone’s description of stereoscopic depth with paired disparate afterimages. According to Brücke, the variations in convergence with successive fixations built up the impression of depth and singleness.This suggestion should have been demolished by Dove’s (1841) demonstration of stereoscopic depth © Japanese Psychological Association 2012.
with brief illumination of stereopairs by means of an electric spark. But it lingered, because without eye movements Brücke and others experienced diplopia with depth perception when viewing one of Wheatstone’s stereograms that had two different sizes of circles in each eye (figure 16 in Figure 5 of this paper). This observation was taken as countering Wheatstone’s (1838) conclusion that “. . . objects whose pictures do not fall on corresponding points of the two retinæ may still appear single” (p. 384). The quantification of binocular disparity by Panum (1858/1940) led to the concept of Panum’s area and it should have discouraged the all-or-none approach in describing the single vision of disparate stimuli, but it continued with LeConte until 1897. He did not know of Panum’s study and repeated Dove’s experiment using a stimulus with a large disparity and observing diplopia with depth perception. Helmholtz (1867) used Dove’s (1841) technique to support Wheatstone’s observations with an improved version of the Wheatstone Experiment (involving two lines presented to each eye). In addition, Dove’s method of brief electrical illumination provided the spark of inspiration for the tachistoscope, which presents visual stimuli very briefly. It was the problems associated with electric sparks that resulted in the search for alternative methods of brief presentation for controlled durations. Volkmann (1859) gave the name tachistoscope to an instrument of his invention, and it was developed initially for observing stereoscopic images. He confirmed Dove’s observation of stereoscopic depth without eye movements. Many different versions of stereoscopes were patented in the second half of the 19th century (Gill, 1969), but most of them were minor variations on the mirror, lens, or prism models. Helmholtz (1857) described his telestereoscope in 1857; it enhanced disparities by extending the separation between mirrors. The anaglyph method, enabling overprinted red and green images to be combined through similarly colored filters, was introduced at about the same time (d’Almeida, 1858), although others were producing similar systems (Blundell, 2011).
Binocular and stereoscopic vision
Wheatstone (1838) had employed outline figures for his stereoscope in order to reduce any monocular cues to depth, but he was acutely aware that some remained. The initial attempts to use stereoscopic techniques to conceal and then reveal images came from an unlikely source: the microanatomist Ramón y Cajal in the 1870s. His interests in photography led him to use stereophotography as a technique for transmitting secret messages (Bergua & Skrandies, 2000). The technique he devised (Figure 8) is not unlike the principle of random dot stereograms developed in the twentieth century (see Blundell, 2011). Cajal (1901) described it thus: During my stereoscopic honeymoon, that is to say, long ago between the years 70 and 72, I was absorbed in imagining new fancies and recreations of this genre. My aim was to achieve a mysterious writing, which could only be deciphered with the stereoscope and usable for those people who don’t want to divulge their own matters. . . . The game consists of making a proof [a print on glass] only with dots, lines and scribbles, or also of letters, crossed and entangled in a thousand ways. A proof in which, with the naked eye, you cannot read anything at all. And, never-
67
theless, as soon as you see the double image of this background in the stereoscope, a perfect legible sentence or text suddenly appears, standing out on the foreground and clearly detaching itself from the chaos of the lines or dots. (Translated by Bergua & Skrandies, 2000, p. 71) Handmade stereoscopic dot patterns were produced by Herbert Mobbs in 1919, by Boris Kompaneysky in 1939, and more complex versions were made by Claus Aschenbrenner in 1954 (Blundell, 2011; Howard & Rogers, 2002). Julesz (1960) enlisted the power of the computer to produce them with greater ease and further developments resulted in autostereograms. Presenting regular dot patterns that enable fusion of neighboring pairs is the basis of the wallpaper illusion, which was described by Blagden (1813) before the stereoscope was invented. When equivalent but laterally separated patterns are combined binocularly they seem suspended in the plane of convergence. Blagden observed the effect by chance when viewing the pattern in a marble chimney-piece, and it was rediscovered by Brewster (1844b) with regularly patterned wallpaper. Another stereoscopic technique involving repetitive patterns was introduced in an artistic
Figure 8 Cajal’s (1901) technique for encoding messages stereoscopically: “Two things are necessary to perform this fantasy: the background with dots, lines, letters or entangled scribbles; and a big clean glass, where you write what you want to stand out with the stereoscope. For the illusion to be perfect, it is necessary that the thickness of the lines or the dots of what you write, be the same when you look through the scraped glass as the one of the lines, letters or dots drawn on the background.” (translated by Bergua & Skrandies, 2000, p. 71). © Japanese Psychological Association 2012.
68
N. J. Wade and H. Ono
context by Ludwig Wilding (1927–2010). It bears some resemblance to Cajal’s use of transparent surfaces and is based on disparities between moiré fringes generated by the interference of regular repetitive patterns (like gratings) separated slightly in depth. Moreover, the apparent stereoscopic space varies with the viewing distance of the observer (because disparity between the moiré fringes varies with viewing distance). The depth can be produced from curved as well as flat surfaces, and opposite directions of depth are often incorporated in the same work (Wade, 2007). The relationship between the spatial frequencies of the transparent and printed patterns determines the direction and amount of the depth seen, and they can be given precise mathematical descriptions (Kondo, Wade, & Nakamizo, 1990). Depth can be seen as a consequence of the disparities of the moiré fringes when the head is stationary, and it can be augmented by lateral head movements that yield motion parallax between the moiré patterns.
ity was considered to introduce a problem in interpreting binocular single vision rather than a partial solution to the perception of depth or distance. The stereoscope transformed not only the vision of pictures, but also the understanding of some intricacies of spatial perception. Wheatstone showed how the photographic camera, in combination with the stereoscope, could be employed to reintroduce the dimension of depth to the perception of pictures. He achieved the result to the problem that Leonardo had struggled so long with. Wheatstone’s obituary notice, published in Nature, contained the following comments of Signor Volpicelli (1876) of the Academia dei Lincei:
Conclusion
Thus, Wheatstone’s invention both solved an ancient puzzle and provided the principles that have been the source of study ever since.
Color vision and space perception are fundamental features of our interaction with the world. The experimental study of both domains was advanced by optical instruments that enabled stimulus control, so that phenomena could be investigated in the laboratory. Great strides were made when such stimulus control could be harnessed. It is argued that what Newton achieved for color vision with manipulations of the prismatic spectrum, Wheatstone achieved for space perception with the invention of the stereoscope and his experiments with it. The stereoscope is a simple optical device that presents slightly different figures to each eye; if these figures have appropriate horizontal disparities then depth is seen. Paradoxically, knowledge of retinal disparity has a history stretching back to Ptolemy and Galen, but the use to which it was put was only appreciated with Wheatstone’s (1838) invention of the stereoscope. That is, the existence of retinal dispar© Japanese Psychological Association 2012.
Our countryman, Leonardo da Vinci, in 1500, or thereabouts, conceived and was the first to affirm, that from a picture it was not possible to obtain the effect of relief. But Wheatstone, reflecting profoundly in 1838, on the physiology of vision, invented the catoptric stereoscope, with which he philosophically solved the problem of the optical and virtual production of relief. (p. 502)
References Aguilonius, F. (1613). Opticorum libri sex. Philosophis juxta ac mathematicis utiles. Antwerp: Moreti. Alberti, L. B. (1966). On painting. (Trans. by J. R. Spencer), New Haven: Yale University Press. (Original work published (1435). Della pittura. Florence.). Alhazen (1572). Opticae thesaurus. (Trans. and Ed. by F. Risner), Basel: Gallia. d’Almeida, J.-C. (1858). Nouvel appareil stéréoscopique. Comptes Rendus Hebdomadaires des Séances de l’Académie des Sciences, 47, 61–63. Arnold, H. J. P. (1977). William Henry Fox Talbot. Pioneer of photography and man of science. London: Hutchinson Benham. Bergua, A., & Skrandies, W. (2000). An early antecedent to modern random dot stereograms – “The secret stereoscopic writing” of Ramón y Cajal. International Journal of Psychophysiology, 36, 69–72.
Binocular and stereoscopic vision Berkeley, G. (1709). An essay towards a new theory of vision. Dublin: Pepyat. Blagden, C. (1813). An appendix to Mr. Ware’s paper on vision. Philosophical Transactions of the Royal Society, 103, 110–113. Blundell, B. (2011). 3D displays and spatial interaction: Exploring the science, art, evolution and use of 3D technologies. Auckland, NZ: Walker & Wood. Boyle, R. (1688). A disquisition about final causes of natural things: Wherein it is inquir’d, whether, and (if at all) with what cautions, a naturalist should admit them? To which is subjoyn’d, by way of Appendix, some uncommon observations about vitiated sight. London: Taylor. Brewster, D. (1830). Optics. Edinburgh encyclopædia. Vol. 15. Edinburgh: Blackwoods, pp. 460–662. Brewster, D. (1844a). On the law of visible position in single and binocular vision, and on the representation of solid figures by the union of dissimilar plane pictures on the retina. Transactions of the Royal Society of Edinburgh, 15, 349–368. Brewster, D. (1844b). On the knowledge of distance given by binocular vision. Transactions of the Royal Society of Edinburgh, 15, 663–674. Brewster, D. (1851). Description of several new and simple stereoscopes for exhibiting, as solids, one or more representations of them on a plane. Transactions of the Royal Scottish Society of Arts, 3, 247–259. Brewster, D. (1856). The stereoscope. Its history, theory, and construction. London: John Murray. Brücke, E. (1841). Ueber die stereoskopischen Erscheinungen und Wheatstone’s Angriff auf die Lehre von den identischen Stellen der Netzhäute. Archiv für Anatomie, Physiologie und wissenschaftliche Medicin, 5, 459–476. Cajal, S. R. y. (1901). Recreaciones estereocópicas y binoculares. La Fotografia (Madrid), 27, 41– 48. Crone, R. A. (1992). The history of stereoscopy. Documenta Ophthalmologica, 81, 1–16. Desaguliers, J. T. (1716). A plain and easy experiment to confirm Sir Isaac Newton’s doctrine of the different refrangibility of the rays of light. Philosophical Transactions of the Royal Society, 34, 448–452. Descartes, R. (1965). Discourse on method, optics, geometry, and meteorology. (Trans. by P. J. Olscamp), Indianapolis, IN: Bobbs-Merrill. (Original work published (1637). Discours de la méthode. Leyden: Maire.). Descartes, R. (1972). Treatise of man. René Descartes. (Trans. & Ed. by T. S. Hall), Cambridge, MA: Harvard University Press. (Original work published (1664). L’homme. Paris: Angot.)
69
Dove, H. W. (1841). Die Combination der Eindrücke beider Ohren und beider Augen zu einem Eindruck. Monatsberichte der Berliner preussische Akademie der Wissenschaften, 41, 251–252. Dove, H. W. (1851). Beschreibung mehrerer Prismenstereoskope und eines einfachen Spiegelstereoskops. Annalen der Physik und Chemie, 83, 183–189. Du Tour, E.-F. (1760). Discussion d’une question d’optique. Académie des Sciences. Mémoires de Mathématique et de Physique Présentés pars Divers Savants, 3, 514–530. Du Tour, E.-F. (1761). Addition au mémoire intitulé. Discussion d’une question d’optique. Académie des Sciences. Mémoires de Mathématique et de Physique Présentés pars Divers Savants, 4, 499– 511. Euclid (1895). Optica. (Ed. by J. L. Heiberg), Leipzig: Teubner. Gill, A. T. (1969). Early stereoscopes. Photographic Journal, 109, 641–651. Helmholtz, H. (1857). Das Telestereoskop. Annalen der Physik und Chemie, 101, 494–496. Helmholtz, H. (1867). Allgemeine Encyklopädie der Physik. Vol. 9. Handbuch der physiologischen Optik. (Ed. by G. Karsten), Leipzig: Voss. Hering, E. (1862). Beiträge zur Physiologie. Vol. 2. Leipzig: Engelmann. Howard, I. P. (2002). Seeing in depth. Vol. 1. Basic mechanisms. Toronto, Canada: Porteus. Howard, I. P., & Rogers, B. (2002). Seeing in depth. Vol. 2. Depth perception. Toronto, Canada: Porteus. James, W. (1890). Principles of psychology. Vol. 2. London: Macmillan. Julesz, B. (1960). Binocular depth perception of computer generated patterns. Bell System Technical Journal, 39, 1125–1162. Klooswijk, A. I. J. (1991). The first stereo photo. Stereo World May/June, 6–11. Kondo, M., Wade, N. J., & Nakamizo, S. (1990). Geometrical analysis of the motion and depth seen in moiré patterns. Bulletin of the Faculty of Literature, Kitakyushu University, 22, 97–114. de La Hire, P. (1694). Dissertation sur les differens accidens de la vue. In F. Boulduc (Ed.), Mémoires de Mathématique et de Physique. Paris: Anisson, pp. 233–302. Le Clerc, S. (1679). Discours touchant de point de veue, dans lequel il es prouvé que les chose qu’on voit distinctement, ne sont veues que d’un oeil. Paris: Jolly. Le Clerc, S. (1712). Système de la vision. Paris: Delaulne. LeConte, J. (1881). Sight: An exposition of the principles of monocular and binocular vision. New York: Appleton. © Japanese Psychological Association 2012.
70
N. J. Wade and H. Ono
LeConte, J. (1897). Sight: An exposition of the principles of monocular and binocular vision. 2nd ed. New York: Appleton. Mayo, H. (1833). Outlines of human physiology. 3rd ed. London: Burgess and Hill. Newton, I. (1704). Optics: Or, a treatise of the reflections, refractions, inflections and colours of light. London: Smith and Walford. Ono, H., Lillakas, L., & Wade, N. J. (2007). Seeing double and depth with Wheatstone’s stereograms. Perception, 36, 1611–1623. Ono, H., & Wade, N. J. (1985). Resolving discrepant results of the Wheatstone Experiment. Psychological Research, 47, 135–142. Ono, H., & Wade, N. J. (2005). Depth and motion in historical descriptions of motion parallax. Perception, 34, 1263–1273. Ono, H., & Wade, N. J. (2012). Two historical strands in studying visual direction. Japanese Psychological Research, 54, 71–88. Panum, P. L. (1940). Physiological investigations concerning vision with two eyes. (Trans. by C. Hubscher), Hanover, NH: Dartmouth Eye Institute. (Original work published (1858). Physiologische Untersuchungen über das Sehen mit Zwei Augen. Kiel: Schwerssche Buchhandlung.) Porta, J. B. (1593). De refractione. Optices parte. Libri novem. Naples: Salviani. Prévost, A. P. (1843). Essai sur la théorie de la vision binoculaire. Geneva. Rohault, J. (1671). Traité de physique. Paris: Savreux. Ross, W. D. (1931). The works of Aristotle. Vol. 3. Oxford: Clarendon. Sabra, A. I. (1989). The optics of Ibn al-Haytham. Books I–III. On direct vision. London: The Warburg Institute. Smith, A. M. (1996). Ptolemy’s theory of visual perception: An English translation of the Optics with introduction and commentary. Philadelphia, PA: The American Philosophical Society. Towne, J. (1862). Remarks on the stereoscopic theory of vision; with observations on the experiments of Professor Wheatstone. Section I. Guy’s Hospital Reports, 8, 70–80. Towne, J. (1869). Contributions to the physiology of binocular vision. Section VIII. Guy’s Hospital Reports, 14, 54–83.
© Japanese Psychological Association 2012.
Volkmann, A. W. (1859). Das Tachistoscop, ein Instrument, welches bei Untersuchung des momentanen Sehens den Gebrauch des elektischen Funkens ersetz. Berichte über die Verhandlungen der königlichen sächsischen 2, 90–98. Volpicelli, P. (1876). Scientific worthies. Nature, 13, 501–503. Wade, N. J. (1983). Brewster and Wheatstone on vision. London: Academic Press. Wade, N. J. (1987). On the late invention of the stereoscope. Perception, 16, 785–818. Wade, N. J. (1998). A natural history of vision. Cambridge, MA: MIT Press. Wade, N. J. (2004). Philosophical instruments and toys: Optical devices extending the art of seeing. Journal of the History of the Neurosciences, 13, 102–124. Wade, N. J. (2007). The stereoscopic art of Ludwig Wilding. Perception, 36, 479–482. Wade, N. J., Ono, H., & Lillakas, L. (2001). Leonardo da Vinci’s struggles with representations of reality. Leonardo, 34, 231–235. Wheatstone, C. (1838). Contributions to the physiology of vision – Part the first. On some remarkable, and hitherto unobserved, phenomena of binocular vision. Philosophical Transactions of the Royal Society, 128, 371–394. Wheatstone, C. (1839). Ueber das Sehen mit zwei Augen und das Stereoskop. Annalen der Physik und Chemie, 47, 625–627. Wheatstone, C. (1842). Beiträge zur Physiologie des Gesichtssinnes. Erster Teil. Annalen der Physik und Chemie, Supplementary, 1, 1–48. Wheatstone, C. (1852). Contributions to the physiology of vision – Part the second. On some remarkable, and hitherto unobserved, phenomena of binocular vision. Philosophical Transactions of the Royal Society, 142, 1–17. Young, T. (1807). A course of lectures on natural philosophy and the mechanical arts. London: Johnson. (Received April 18, 2011; accepted July 2, 2011)
