The astonishing images created by the Old Masters have been attributed to two causes - either their extraordinary artistic skill, using eye, hand and pencil, or through the use of mathematical perspective, developed during the Renaissance in the early 1440s. Over the last few years, however, David Hockney and his collaborator, the physicist Charles Falco, have been exploring a third possibility...
Hockney and Falco consider this 1543 double portrait by Lorenzo Lotto to be a striking demonstration of their theory. Look at the oriental carpet lying on top of the table. The distant edge of the keyhole pattern is distorted and out of focus. This pattern could not have been sketched: when we look at something, it is never out of focus. Instead, optics were used to record this complex carpet pattern. An optical image of the carpet was projected onto the surface of the painting and the image was traced. But only part of the optical image would have been in focus. To trace the rest, the image had to be moved and refocused. (Note that the more distant carpet pattern is back in focus again.) Furthermore, we can identify the number of vanishing points in the picture. We find two of them, as expected for the re-focusing of an optical image. Simple "eyeballing" and sketching would only have yielded one. Use of geometrical perspective would also have yielded one. Lotto seems to have traced an optical image.
When and where were optical projections first used? When and where do paintings first show an "optical look", that suggest the use of optical techniques. Consider this sequence of four paintings, ranging from 1300 to 1430. The first three are not lifelike, they lack characteristic, detailed human features. The fourth, painted in Bruges around 1430, is an individual we would be able to recognize, almost a photographic image. Suddenly, in Bruges around 1430, something happened in Western painting. Awkwardness disappeared: realistic human beings appeared. How did this happen?
Cardinal Niccolo Albergati visited Bruges just once, for three days in 1431. Two portraits of him, by van Eyck, date from this visit - a drawing (half lifesize) made at the time, and a painting (40% larger than the drawing), made a year later. The drawing seems to be a tracing of an optical projection: pinprick eyes and dark shadows suggest the use of bright, outdoor lighting. The drawing and the later painting are effectively superimposable. The painting was traced from an enlarged optical image of the drawing, projected with a mirror. Rapid, detailed sketching and accurate enlarging and copying suggest that van Eyck used optics.
Optical images can be traced using lenses or mirrors. Lenses were not available in Bruges in the 1430s. But mirrors were, as van Eyck's painting here shows; and Charles Falco, a professor of optics, has emphasized that lenses and concave mirrors have similar focusing properties. (Use a shaving mirror to focus a bright exterior image onto a dark interior wall: the image is upside down and traceable.) van Eyck's painting shows a convex mirror on the wall: the convex mirror becomes a concave mirror if it is turned around and the silvering is reversed. Optical devices were available in Bruges in the 1430s for van Eyck to use.
Hockney first identified the use of optical techniques, at the major Ingres retrospective in London in January 1999. Hockney was dumbstruck by the photographic clarity and precision of the small pencil portraits, drawn quickly by Ingres, of unfamiliar visitors to Rome around 1820. Several features of these drawings suggested that Ingres had been using a camera lucida - essentially a prism on a stick, that projects an image that the artist can trace. Blown up, details of these portraits resembled details of Andy Warhol's tracings of images produced by a slide projector. There is a sureness of line without uncertainty, suggesting the following of an existing outline. Immediately Hockney began to see the same "optical look" in the work of Caravaggio, Holbein, Vermeer and Velazquez.
When a complicated scene is painted on a large canvas, an image of the scene can still be projected onto the canvas and traced, but problems arise. Only part of the image will be in focus on the canvas and the lens must be moved and refocused several times to trace the whole scene. The refocusing introduces multiple vanishing points and anatomical distortion. Three examples are shown. In the painting by Caravaggio, St. Peter's outstretched right hand (farthest away) is larger than his left hand and Jesus' right hand (which are closer). In the van Dyck, the lady's lower limbs dwarf her upper limbs. Again, in the Chardin, the maid seems to have two elbows and too small a head. These distortions support Hockney's argument that optical techniques were used.
Hockney and Falco maintain that the "optical look", the product of "optical devices", influenced Western art from 1430 to about 1850, when the development of chemical fixatives established photography. Henceforward, photography, not painting, would render "reality". Painting rejected the "optical look" and, through movements, such as Impressionism and Cubism, etc., critiqued photography, capturing facets that a mere photograph could not - time, multiple vantage points, emotion, lived reality rather than reality. European painting regained awkwardness again, for the first time in over 400 years. The pre-optical look of the Byzantine Christ (c. 1150) links to the post-optical look of the van Gogh portrait (1889).