Vermeer and the Camera Obscura: Part III

The limitations of the camera obscura as an aid to painting

Depth of field. One of the most limiting factors in utilizing the camera obscura as an aid to painting, which is particularly crucial for tracing, is its extremely narrow depth of field (fig. 1), an inexorable product of a single convex lens configuration. Depth of field is defined as the distance between the nearest and the farthest objects in an image that are in acceptably sharp focus (in reality, a lens can only focus sharply at one point, but the transition from sharp to unsharp is gradual, and thus the term "acceptably sharp"). The field of depth attainable with a single-lens optical apparatus can be broadened by restricting the diameter of the lens or by distancing the lens from the subject matter. But it is not due to the inferior optics of the seventeenth century as been sometimes supposed. As experiments show, with such a large projection size in the camera, refocusing is required even with a modern, corrected triplet objective.^{John Smith, The History of Time (New York: Random House, 2000), 45.} Another limitation of camera obscura as an aid to painting is the dimness of its projection.

fig. 1 A photograph taken from the screen of a portable camera obscura showing a scale mockup of Vermeer's A Lady Standing at the Virginals, which is focused on the figure. While the figure is acceptably in focus notice how blurry the black-framed Cupid paining hanging on the background wall only a few meters behind the figure.

As an example of the first limitation, had the lens of Vermeer's camera been focused on the background wall of scene represented in The Art of Painting, the foreground chair would have been far too blurry to make a precise tracing. Moreover, in those works where the floor occupies a prominent part of the composition, it is difficult to discern, and hence trace with requisite precision, the floor tiles, whose contours are of fundamental importance for creating mathematically accurate perspectives. If the tiles that abut the background wall were brought into tight focus, then those nearest the viewer—i.e., those along the lower edge of the painting—would be so blurry and dim that tracing them would involve a significant dose of guesswork. To resolve this dilemma it would seem that one has only to refocus the lens and continue tracing. However, refocusing is accomplished by moving the lens closer or farther from its initial position. This movement alters the dimensions of the projected image so that the contours of the tiles of the two tracings no longer precisely connect: one set is slightly larger than the other. Moreover, by moving the position of the lens the perspective is altered because the scene is viewed from a slightly different position (point of view). Refocusing and maintaining one central perspective are thus mutually exclusive. It has been sometimes suggested that refocusing can be avoided by simply stopping down (restricting) the aperture diameter of the lens, but in practice this causes the already dim projection to become dimmer still, making it virtually impossible to distinguish the outlines of objects that are not intensely illuminated. Even if Vermeer had discovered a way to bring the whole image into focus at one instant—there is no evidence that seventeenth-century lens makers had developed the technology for overcoming this decisive limit—precision tracing within the poorly lit parts of the scene would remain problematic.

In 2006, the painter Carsten Wirth and Thiemen Coquyt attempted to apply our current knowledge of seventeenth-century optics to the camera obscura in a studio setting in the Utrecht University Museum. After experimenting with various types of historical and modern lenses Coquyt concluded that "compromises between image brightness and depth of field can be obtained."^{Susan Johnson, "The Space Race: A Modern Perspective," Astronomy Today 15, no. 3 (2017): 28-29.} Unfortunately the experiments were not carried out in environmental conditions analogous to those in Vermeer's studio, although the observations of Wirth and Coquyt are particularly illuminating, given the lack of any experimetally valid testing of the camera obscura.

Luminosity. It is true that a wider lens—Steadman advances that Vermeer could have had a lens from 7 to 10 centimeters in diameter but others speculate much bigger lenses would have been available—produces a more luminous image, but with a proportionate loss of image sharpness. In fact, there are many parts of Vermeer's interiors that are so scarcely illuminated, such as the foregrounds of The Concert, Woman with a Pearl Necklace or Woman Holding a Balance, that the outlines of single objects would have sunk into an indistinct, milky darkness and would have been impossible distinguish, much less trace. In some cases, this handicap can be remedied by simple workarounds that could have been devised by any reasonably resourceful seventeenth-century painter. For example, Vermeer could have temporarily opened all (three) of his studio's windows and removed any curtains that had been used to "shape" the illumination, perhaps on a particularly sunny day, remembering that the texture, color and shape of shadows are irrelevant to perspective and outline drawing: all that is needed are edges. In certain circumstances it would also have been possible to reflect the incoming light back into the shadowed side of a particular object by holding a large, properly angled white board behind it. This is similar to how today's photographers use so-called "flats" or "bounce boards," especially with interior of buildings lacking sufficient available light. But even if enough light could be produced by one or another trick, the question of field of depth remains.

Compositon. Although at first glance the camera obscura might seem a near perfect "composition machine" its limited focal range and left-to-right reversal are a limiting factor in fine tuning compositional relationships. In effect, although camera's image is undoubtedly "pictorial," it is difficult to see how objects located on different focal planes could be weighed and gauged reciprocally such that they would allow the level of precision of compositonal alignment and juxtapositioning typical of Vermeer's designs. Moreover, the process of revising the positions of the actual objects outside the camera obscura in order to obtain more precise two-dimensional compositonal dispositions on the camera's dim and unvenly focused screen is hindered by the fact that one must allow some time to elapse when entering the camera obscura booth so that the eyes can readjust to the camera's dim image.

Image reversal. As has been said, the image projected by the portable camera obscura is upside-down and mirror-reversed. This does not effect tracing but produces problems for making direct visual comparisons with the real scene. By placing a mirror at a 45 degree angle inside the camera with respect to the direction of the light coming from the lens, the image can be uprighted, but due to the laws of optics, it remains left-to-right reversed. However, once the image had been traced, there existed various well-known and efficient methods that would have allowed a trained artist to rectify the left-to-right reversal before transferring it to his canvas so it would have the same orientation as that of the original scene. But, there seems to be no practical means to correct both defects in orientation of image in the portable camera.

Whether or not the camera can be effectively utilized as a composition machine, the difficulties of orientation, which may appear rather significant for the layman, are not insurmountable. They are, in fact, less tricky for the practicing painter whose job it is to solve visual problems of far greater technical complexity.

In order to overcome the camera's faulty orientation it has been suggested that Vermeer "built" the actual studio setup mirror-reversed so that the camera's image would assume the orientation we see in his pictures. He would have used an environment with the windows to the right instead of to the left, and placed the objects we see on the right-hand side of his pictures to the left-hand side of the real scene. The figures could have easily held objects like wineglasses or stringed instruments in their left hands, necessitating only minor adjustments in drawing. Obviously, completely symmetrical objects would present no problem in that they would appear identical in both the original and reversed image. The problem arises when dealing with pictures-within-pictures and, above all, the extraordinarily intricate designs of the wall maps. Steadman remarked that Vermeer could have made his initial outline drawing directly from the camera's image leaving blank outlines in place of the maps or pictures-within-pictures. "He would then have made a separate tracing on paper of the image of each map or painting, reversed this tracing, and reintroduced it in the appropriate position in the image of the whole scene. Because all of these items are seen frontally, and their images are therefore simple rectangles, this would have been a perfectly feasible and straightforward procedure."

Reversing the room would have allowed the painter to work almost integrally from the image projected by camera's lens. But, obviously, he would have had difficulty when consulting the real room when he needed to apply color. Steadman pointed out that it is known that professional photographers who once used plate cameras became accustomed to assessing compositional relationships in inverted images, and that, curiously, the problem might have had the virtue of forcing Vermeer to see the composition more abstractly in terms of relations between shapes.

Detail. Some have noted that if the camera's lens is placed at the distance from which the Vermeer observed his scenes, the details of the wall-maps would have been barely, if at all, discernible. This is true. But it would be a simple matter to temporarily hang the map near an open window and move the camera close to the map so that it would fill the whole screen. In this manner the finest details of the map could be traced with satisfactory precision. This drawing could be reversed, resized and finally, as suggested by Steadman, reintroduced to the appropriate rectangle on the canvas.

Are the difficulties discussed herein such that they cannot be overcome by a resourceful artist, especially a highly skilled artist of the past such as Vermeer? It is important to remember that, in fact, nothing obliges an artist using the camera obscura to produce a finished and completely accurate drawing from start to the end. Too many arguments leveled against the camera obscura as a tracing machine assume by default that any departure from an integral and exclusive utilization of the camera during the painting process somehow constitutes proof that the device could not have been used at all. It is all to obvious to any practicing painter that even a reasonably resolved drawing may be of great benefit in as much as it can be reworked to a much higher level of precision once the artist has removed it from the apparatus and set it in more comfortable circumstances. In fact, in the opinion of this writer, rather than a dogmatic application of a clear-cut and repeatable technical program, the highest forms of advanced realism were unachievable without a continuous recourse to what we might term technical and strategic hacks, largely unknown to the greater public. There seems to be no reason to believe that Vermeer could not have devised the simple hacks described above—and who knows how many more—had he felt that the information offered by camera obscura vision was crucial to his artistic discourse.

After a variety of experiments with the camera obscura it can be established that Vermeer’s depictions of complex interiors can never correspond to one single projection by a historical camera obscura optic. If his pictures actually were produced using a camera obscura, this process involved a specially developed, complex method that allowed him to manipulate projected images to achieve the intended results. Thus multiple projections must have been used to produce a painting, by combining the desired qualities of each single one of them. Accordingly, the search for a working model of a "camera obscura a la Vermeer" does not mean seeking a mirror-image correspondence between one camera projection and the painted picture, but rather for a method that unifies different aspects of various projections in the process of painting to manufacture one picture.

Which kind of camera obscura could Vermeer have used?

The considerations that follow assume that both types of lens-fitted camera obscuras would be utilized in conditions analogous to those represented in Vermeer's paintings.

Two kinds of camera obscuras would have been available to Vermeer: the booth (or cubicle; fig. 2) type and the portable type. The booth type is a sort of tent or closed box equipped with a lens whose interior is spacious enough for an observer to be comfortably seated inside. The principal advantage of the booth type is that because it can be sealed securely, the only light that enters will be the light that passes through the lens aperture. These rays are projected onto a sheet of paper or, in the case one has no need to rectify the image, a canvas that is hung vertically on the booth's rear wall. The whiter the paper or canvas, the easier it is to trace. This is not because one can clearly see the drawn lines, but because the relatively luminous background allows one to position the tip of the drawing tool on the drawing surface with greater precision, a tricky task in itself. Inside the camera the thin outlines of the drawing are barely visible so that one must remember which areas have been previously drawn. Although the projected image is relatively clean and legible, it is never objectively brilliant. Any apparent brightness is owed to the surrounding darkness, somewhat like the old photographic slides on a luminous table. In any case, a period of time must elapse before one's eyes become accustomed to the darkness and enable the observer to perceive the camera's image at its fullest.

fig. 2 A hypothetical view of Vermeer tracing the scene of The Music Lesson inside a cubicle type camera obscura (drawn from three drawings of Philip Steadman, Vermeer's Camera: Uncovering the Truth behind the Masterpieces (2002)

fig. 3 A modern portable box-type camera obscura

The portable type of camera is made of a small box (fig. 3), which may have different dimensions, as small as a shoe box if desired. While the booth-type projects the image on the back inside wall, the box type must have a semi-transparent screen on the back or the top side—if a mirror is used to upright the image—on which to throw the projection, since, obviously, the viewer remains outside the box. If the lens is fixed to the front side, the image cannot be refocused. Refocusing with this configuration can be achieved only by moving the camera farther or closer the objects of interest. In order to permit refocusing without moving the body the camera, it can be fitted with a simple two-piece telescoping mechanism which allows the lens itself to move back and forth.

If a portable device is placed on a table in the proximity of the painter's easel, it can be consulted without interrupting the painting process. While such a device can be carried comfortably under one's arm and permits exploration of any well-lit scene, the surrounding light will spill on the screen and dilute the quality of the image. To be practical as a tool for painters, some sort of device, a rigid cowling or a cloth hood, similar to those used by earlier photographers, is needed to shield the camera's screen.

Neither kind of camera obscura was listed in Vermeer's death inventory. However, it should be remembered that the assembly of each kind, and therefore the disassembly, is uncomplicated. In fact, the construction of a camera obscura is a very simple matter. Anyone with discreet manual skill, a mat knife, masking tape, a cheap magnifying lens, a sheet of tracing paper and bit of cardboard can construct a perfectly functional portable box-type camera in the space of a morning. To construct a large booth-type camera, one needs only a greater amount of material to size up the box, and fit it with a door and a seat inside in order to remain comfortably seated while tracing (it is best to paint the interiors of both device's with opaque black paint in order to minimize unwanted reflections). In Vermeer's time, the only component of the camera that could not have been made by an amateur was the lens, which has to be hand ground by someone expert in the field.

Painting in the box?

"There is no argument among today’s historians that the optical camera obscura, which emerged in the final third of the sixteenth century, exerted an influence of some kind on the aesthetics of a number of pictures of the seventeenth century. Yet how a practical and conceptual application of the camera obscura might have looked in concrete terms is still unclear today."^{John Smith, The History of Time (New York: Random House, 2000), 47.} It is true that one can produce a drawing inside a booth-type camera at first go, but to create a drawing useful for advanced forms of illusionist painting considerable practice as well as a period of psychological adaptation. There have been repeated suggestions that Vermeer actually painted in the box but there is no credible evidence in favor of this hypothesis. The penetrating odor of turpentine and essential oils necessary for painting would become quickly unbearable in even the largest booth camera. More importantly, it is impossible to mix colors. The interior of the camera is so dim that the colors on the palette can barely be distinguished, which makes the modulation of tonal values necessary for fine painting hopelessly impracticable. Moreover, when one applies paint to the canvas the lens continues to project colored light. This causes the painted areas to glare and look darker and more colorful than the original projection. The combination of these defects quickly renders the projected image unusable. Thus far, of those who have hypothesized methods to paint directly from the camera's image, none have proven to be remotely practical.

To date, the experiments with the camera obscura that have been carried out by ernest practicing painters were done in what could be called optimal conditions, albeit none with malicious intentions, and none of them have minimally duplicated the conditions under which Vermeer would have worked. Still life painters naturally group their objects tightly on the same plane thereby mitigating the thorny problem associated with the device–s narrow field of depth, and often the subject matter was illuminated with strong constant, artificial light. A few experiments have even used flat prints of Vermeer's paintings as subject matter. This approach completely bypasses any of the real limitations imposed by the camera obscura and creates an erroneous impression of a "Vermeer-like" result.

In any case, those who maintain that Vermeer must have devised a method for painting directly from the screen fail to take into account the skills and near monstrous visual memories possessed by so many Dutch painters, as well as the ability to manipulate their crude materials in unexpected manners. Armed with little more than a collection of impressions, schemata and hastily drawn sketches, it can only be guessed how Dutch seascape painters succeeded in "halting" crashing waves, teetering ships, wind-torn clouds and dramatic shifts of light and shadow in the open sea long enough to capture them in even a quick sketch, much less a fully colored oil painting (fig. 4).

fig. 4 Ships in Distress off a Rocky Coast (detail),
Ludolf Backhuysen,
1667, Oil on canvas, 147.3 x 146.1 cm., National Gallery of Art, Washington D.C.

For most of today's museum goers it is almost impossible to fully appreciate the difficulties of evoking the colors, tones and transparencies of mist, water, vaporous clouds and immaterial shafts of light with nothing but unctuous paint. Certainly, in order to emulate the peculiar, albeit predictable effects of the camera obscura while working from dead-still props while comfortably seated in a studio illuminated by constant Northern light, there would be no need for such a highly trained and intelligent painter as Vermeer to paint directly from the camera's projection once he had assimilated the logic of the optical effects that had impressed him most. The difficulties of translating into paint the peculiar, albeit fairly predictable optical effects of the camera which sat only a few steps from his easel are virtually nil when compared to those faced by the Netherlands's first-tier landscape and marine painters.

Thus, the incessant search for optical configurations that might have allowed Vermeer to transpose the image of the camera obscura directly to his canvas appears by and large unwarranted. For although a number of the artist's pictures do indeed exhibit peculiarities typical of the machine's image, there is no compelling visual evidence that he bound himself to an optical dogma that would necessitate mechanical transcription. In fact, when his works are examined systematically, it becomes immediately evident that he used the camera very selectively, fusing the information that it allowed him to gather with what he observed directly. One would imagine that, in the best of cases, he found the device a novel and efficient means for studying certain aspects of composition, and that once he had reached a satisfactory solution, he could have employed it as shortcut for drawing and perspective. And he almost certainly used the device to ponder the subtleties of tone and color with more objectivity than through naked-eye observation alone. But it appears likely that once he assimilated certain specific vocabularies of camera obscura vision that he found most in line with his artistic interests he could have easily developed these into a pictorial style, which could be applied "freehand," so to speak, without the need to consult the machine's image in every instance.

It should be remembered that unlike the telescope and microscope, both of which were being perfected at the time when Vermeer painted, the camera obscura does not extend human perception. It does the opposite. Color is less brilliant, detail is largely obliterated, and the narrow focal length the device's lens allows only a fraction of the total field of depth of a given scene to be observed with any degree of accuracy. Without much light to power it, the world is plunged into a suggestive dimness, stripped of depth and detail. Moreover, it produces various optical aberrations that obfuscate, rather than reveal the nature of reality. Nonetheless, to the scientific thinker the device may suggest how the eye itself functions. But if such knowledge concerned Vermeer it is not clear from his paintings. They, in fact, raccount another story: that of the limits and the peculiarities of the image it produces. In effect, the device was functional to the ends of the painter precisely because it limits visual information narrowing automatically and unintentionally the impossible range of visual phenomena closer to what can be a captured by the crude means of the painter. It shows not so much how things exist but how they appear when violated by a lens and transformed into flat, colored shapes, which is precisely what seems to matter most for the curious picture maker, Vermeer.

At this point it is worth emphasizing that a practical study of the camera obscura as an aid to painting is closely linked to the circumstances in which it is tested. Most of the experiments that critics have chose to consult have little historic, scientific or practical legitimacy. It is relatively easy to stage a portrait or a simple still life near a window on a sunny day and produce a line drawing suitable for developing a painting. In this case both the colors and the contrasts of the camera's projection may be genuinely informative. And if the pictured objects are placed close to each other and on the same plane, and the camera is sufficiently distanced, their contours can be traced with sufficient precision. But for various reasons, as we have seen, it is entirely another matter to produce a drawing suitable for fine painting a room and its contents that display a relatively deep space such as those typical of Vermeer's most ambitions works. Each object of the still life receives a vastly greater amount of light than the great majority of the props represented in Vermeer's pictures. The overall light of The Concert or A Lady Seated at a Virginal is so low that it may be legitimately questioned if the camera would have been able to "see" anything at all, much less allow the artist to devise perspectives as precise as those of these two works without considerable hacks. It may be surprising that while much has great attention has been paid to gathering and analyzing historical evidence, most writers have habitually referenced experiments that in no way replicate the true, physical circumstance in which Vermeer actually made his paintings.

As a final note, there can be no doubt that Vermeer could have painted without a camera obscura. But there is also little doubt that his paintings would not have displayed some of the characteristics, which, perhaps, confer his work its unique poetry that is so highly treasured today but which were ignored until two hundred years after his death. At present, however, it is not possible to definitively establish to what extent he employed the device. Perhaps, systematic experiments conducted in a real studio setting by qualified artists might disclose something more.

If Vermeer indeed used the camera as an integral part his painting procedure, it is likely that would have traced the most significant outlines of his composition from the camera obscura, and then, moving back and forth, studied the optical images in the darkness, painting on the canvas in full light. He would have relied, so to say, on his visual memory to carry the tonal information acquired from the optical image to his canvas.

Did Vermeer hide his camera obscura?

We do not know if Vermeer made public his use of the camera obscura. Neither a complete apparatus nor lenses of any sort were reported in an inventory of movable goods taken in the artist's house and studio immediately following his death. The pointillés which appear on Metsu's emulation of Vermeer (Woman Reading a Letter, c. 1664–1666) do not prove that he knew how they were produced.

However, the camera obscura had been openly publicized and recommended by contemporary art connoisseurs. So it would seem that there was really no use for hiding it, especially because only those of the small but educated circle were familiar with the device's peculiarities and would have noted them if they saw them in a painting. And yet, there is at least one documented case in which an artist attempted to conceal his working with the camera.

After having bought a camera obscura in London from Cornelis Drebbel in 1622, Constantijn Huygens returned to the Netherlands and enthusiastically demonstrated his new camera to his painter friends. One day the enigmatic Dutch painter Torrentius (fig. 5) "accompanied by several men of standing" urgently asked to see him. Huygens, accompanied by his painter friend Jacques de Gheyn (I), demonstrated the device to the painter and raccounted the event as so:

Torrentius, who displayed his usual submissive modesty and polite manners, looked with feigned amazement at the dancing figures and asked if these little people he saw were actually present as living creatures outside the camera. I confirmed this. As soon as my friends left I remembered his naïve question and his feigned ignorance concerning something everyone knows about these days. I suspected that he was very well aware of the invention but had wanted to create the impression that he was not.

Thus, it would seem that the "cunning fox," as Huygens had called Torrentius, must have already known the device which Huygens thought had lent a "certain a quality to his paintings which the general run of people ascribe to divine inspiration." By keeping his secret the painter may have wished to explain why his still lifes were so realistically painted while his nudes and figure compositions were so dreadfully painted.^{Susan Johnson, "The Space Race: A Modern Perspective," Astronomy Today 15, no. 3 (2017): 30.}

fig. 5 Emblematic Still Life with Flagon, Glass, Jug and Bridle
Torrentius (Johannes van der Beeck)
1614
Oil on panel, 52 x 50.5 cm.
Rijksmuseum, Amsterdam